""" ÃܶȹÀ¼Æ """ import cv2 import json import random, timm from configs.configs import model_head from tqdm import tqdm import os, sys, shutil import torch from torch import nn from PIL import Image, ImageDraw, ImageFont from vizer.draw import draw_boxes # from ssd.config import cfg # from ssd.data.datasets import VOCDataset import numpy as np # from ssd.data.transforms import build_transforms # from ssd.modeling.detector import build_detection_model # from ssd.utils.checkpoint import CheckPointer from fine.models.LoadModel import MainModel from fine.fineConfig import LoadConfig, load_data_transformers from torch.autograd import Variable from creat_xml_fordetection import make_xml import threading from common import global_para as gp import time, statistics from numpy import random root_path = os.getcwd() sys.path.insert(0,root_path+"/Y5") from Y5.models.experimental import attempt_load from Y5.utils.augmentations import letterbox from Y5.utils.general import non_max_suppression, scale_coords from Y5_old.utils.plots import plot_one_box from network_multi import Multi_transform_image from cv2AddChineseText import cv2AddChineseText class Density: def __init__(self, device, root): self.device = device self.root = root self.net = PKSNet().to(self.device) self.pth = "../data_MPKFNet/test/1.pth" # print("creat model is ", (t4 - t3) * 1000) # self.checkpointer = CheckPointer(self.net) # self.net.load_state_dict(self.checkpointer['net']) # self.checkpointer.load(self.checkpointer['net']) checkpoint = torch.load(self.pth) self.net.load_state_dict(checkpoint['net']) if gp.DebugPrint_flag: print("load model weight ok") np.set_printoptions(threshold=np.inf) def inference(self,image,imagetime,index): rootpath = os.path.join(self.root, imagetime) if not os.path.exists(rootpath): os.mkdir(rootpath) if not os.path.exists(rootpath + "/" + "Density"+"/"): os.mkdir(rootpath + "/" + "Density"+"/") target_h, target_w = image.shape[0], image.shape[1] # t1 = time.time() image = cv2.resize(image, (1024, 768)) / 255.0 image = np.transpose(image, [2, 0, 1]) image = np.expand_dims(image, axis=0) image = torch.tensor(image) image = image.to(self.device) # t2 = time.time() # if gp.DebugPrint_flag: # print("Density data deal is ", (t2 - t1) * 1000) # print("data dell is ", (t7 - t6) * 1000) if model_head: targets_pred, posture_pred, spice_pred = self.net(image) else: targets_pred = self.net(image) # t3 = time.time() # if gp.DebugPrint_flag: # print("Density inference is ", (t3 - t2) * 1000) targets_pred1 = targets_pred.cpu() num = targets_pred1.data.sum().numpy() targets_pred1 = targets_pred1.data.numpy() # t4 = time.time() # if gp.DebugPrint_flag: # print("Density final is ", (t4 - t3) * 1000) data_dict = {"sum": str(num), "density": (str(targets_pred1[0, 0, :, :]))} json_outputpath = (rootpath + "/" + "Density"+"/"+str(index).zfill(3)+".json") json_data = json.dumps(data_dict, ensure_ascii=True, indent=4) with open (json_outputpath,'w',encoding='utf-8') as f: f.write(str(json_data)) # t5=time.time() # if gp.DebugPrint_flag: # print("Density save is ", (t5 - t4) * 1000) def inference_init(self, image): image = cv2.resize(image, (1024, 768)) / 255.0 image = np.transpose(image, [2, 0, 1]) image = np.expand_dims(image, axis=0) image = torch.tensor(image) image = image.to(self.device) if model_head: targets_pred, posture_pred, spice_pred = self.net(image) else: targets_pred = self.net(image) targets_pred1 = targets_pred.cpu() num = targets_pred1.data.sum().numpy() targets_pred1 = targets_pred1.data.numpy() if gp.DebugPrint_flag: print("Density initing") """ Ä¿±ê¼ì²â """ class Detection_S: def __init__(self,device,root): self.device = device self.root = root cfg.merge_from_file("ssd/config/hat-512.yaml") cfg.freeze() self.net =build_detection_model(cfg).to(self.device) self.checkpointer = CheckPointer(self.net, save_dir=cfg.OUTPUT_DIR) self.ckpt = "../data_MPKFNet/weigth/detection/detection.pth" self.checkpointer.load(self.ckpt, use_latest=self.ckpt is None) self.transforms = build_transforms(cfg, is_train=False) self.threshold = gp.detect_thresh self.net.eval() def inference(self,image,imagetime,index,fine_grained): with torch.no_grad(): class_names = VOCDataset.class_names rootpath = os.path.join(self.root, imagetime) if not os.path.exists(rootpath): os.mkdir(rootpath) if not os.path.exists(rootpath + "/" + "Detection" + "/"): os.mkdir(rootpath + "/" + "Detection" + "/") image_or = image height, width = image.shape[:2] # t1 = time.time() image = self.transforms(image)[0].unsqueeze(0) # t2 = time.time() # if gp.DebugPrint_flag: # print("Detection data deal is ",(t2-t1)*1000) result = self.net(image.to(self.device))[0] # t3 = time.time() # if gp.DebugPrint_flag: # print("Detection inference is ",(t3-t2)*1000) result = result.to(torch.device("cpu")) result = result.resize((width, height)).numpy() # t5 = time.time() # if gp.DebugPrint_flag: # print("Detection get result is ", (t5 - t3) * 1000) boxes, labels, scores = result['boxes'], result['labels'], result['scores'] indices_score = (scores > self.threshold) indices_label_left = labels <4 indices_label_right = labels >2 indices = [] for i in range(len(indices_score)): indices.append(indices_score[i] and indices_label_left[i] and indices_label_right[i]) boxes = boxes[indices] labels = labels[indices] scores = scores[indices] for i in range(len(boxes)): if boxes[i][0]<0: boxes[i][0] = 0 if boxes[i][1]<0: boxes[i][1] = 0 if boxes[i][2]<0: boxes[i][2] = 0 if boxes[i][3]<0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 99999 boxes_fine, labels_fine, scores_fine = fine_grained.inference(image_or, boxes, labels, scores) # t4 = time.time() # if gp.DebugPrint_flag: # print("Detection final is ", (t4 - t3) * 1000) #numb = 2 #fine_grained.inference(image_or,boxes,labels,scores) # for box in boxes: # img1 = image_or.copy()[int(box[1]):int(box[3]), int(box[0]):int(box[2]), :] # cv2.imwrite(rootpath + "/" + "Detection" + "/"+"{}.jpg".format(numb), img1) # numb+=1 #drawn_image = draw_boxes(image_or, boxes_fine, labels_fine, scores_fine).astype(np.uint8) xml_outputpath = rootpath + "/" + "Detection"+"/" + index+".xml" make_xml(str(index),image_or.shape[0],image_or.shape[1],boxes_fine,labels_fine,scores_fine,xml_outputpath,class_names) #Image.fromarray(drawn_image).save(xml_outputpath.replace(".xml",".jpg")) # t5 = time.time() # if gp.DebugPrint_flag: # print("Detection save is ", (t5 - t4) * 1000) def inference_Red(self,image,ImagePath,fine_grained): with torch.no_grad(): class_names = VOCDataset.class_names image_or = image height, width = image.shape[:2] # t1 = time.time() image = self.transforms(image)[0].unsqueeze(0) # t2 = time.time() # if gp.DebugPrint_flag: # print("data deal is ",(t2-t1)*1000) result = self.net(image.to(self.device))[0] # t3 = time.time() # if gp.DebugPrint_flag: # print("inference is ",(t3-t2)*1000) result = result.to(torch.device("cpu")) result = result.resize((width, height)).numpy() boxes, labels, scores = result['boxes'], result['labels'], result['scores'] indices_score = (scores > self.threshold) indices_label_left = labels <4 indices_label_right = labels >2 indices = [] for i in range(len(indices_score)): indices.append(indices_score[i] and indices_label_left[i] and indices_label_right[i]) boxes = boxes[indices] for i in range(len(boxes)): if boxes[i][0]<0: boxes[i][0] = 0 if boxes[i][1]<0: boxes[i][1] = 0 if boxes[i][2]<0: boxes[i][2] = 0 if boxes[i][3]<0: boxes[i][3] = 0 labels = labels[indices] scores = scores[indices] for i in range(len(labels)): labels[i] = 99999 boxes_fine, labels_fine, scores_fine = fine_grained.inference(image_or, boxes, labels, scores) # t4 = time.time() # if gp.DebugPrint_flag: # print("final is ", (t4 - t3) * 1000) #numb = 2 #fine_grained.inference(image_or,boxes,labels,scores) # for box in boxes: # img1 = image_or.copy()[int(box[1]):int(box[3]), int(box[0]):int(box[2]), :] # cv2.imwrite(rootpath + "/" + "Detection" + "/"+"{}.jpg".format(numb), img1) # numb+=1 #drawn_image = draw_boxes(image_or, boxes_fine, labels_fine, scores_fine).astype(np.uint8) xml_outputpath = ImagePath.replace("/ftp/", "/ftp_Result/").replace(".JPG", ".xml").replace(".jpg",".xml") make_xml(str(ImagePath),image_or.shape[0],image_or.shape[1],boxes_fine,labels_fine,scores_fine,xml_outputpath) #Image.fromarray(drawn_image).save(xml_outputpath.replace(".xml",".jpg")) # t5 = time.time() # if gp.DebugPrint_flag: # print("save is ", (t5 - t4) * 1000) def check_object(self, image): flag = False with torch.no_grad(): height, width = image.shape[:2] img_temp = image # t1 = time.time() image = self.transforms(image)[0].unsqueeze(0) # t2 = time.time() # if gp.DebugPrint_flag: # print("check datadeal is ",(t2-t1)*1000) result = self.net(image.to(self.device))[0] # t3 = time.time() # if gp.DebugPrint_flag: # print("check inference is ", (t3 - t2) * 1000) result = result.to(torch.device("cpu")) # t4 = time.time() # if gp.DebugPrint_flag: # print("check result deal is ", (t4 - t3) * 1000) result = result.resize((width, height)).numpy() boxes, labels, scores = result['boxes'], result['labels'], result['scores'] indices_score = (scores > self.threshold) indices_label_left = labels <4 indices_label_right = labels >2 indices = [] for i in range(len(indices_score)): indices.append(indices_score[i] and indices_label_left[i] and indices_label_right[i]) labels = labels[indices] boxes = boxes[indices] scores=scores[indices] if 3 in labels: flag = True # if gp.DebugPrint_flag: # print("boxes = ", boxes) # print("check flag is ", flag) # t5 = time.time() # if gp.DebugPrint_flag: # print("check final is ", (t5 - t4) * 1000) return flag, labels, boxes, scores class Detection_Y5: def __init__(self,device,root,detect_thresh): self.device = device self.root = root self.img_size = 2176#640 self.conf_thres = detect_thresh self.iou_thres = 0.4 self.stride = 1 # self.weights = "/data/Hatcher/data/Detection_weigth/BASE_PYH_SJH-Bird-1-45328-2160-s-base9/weights/best.pt" self.weights = gp.DetWeight # self.device = select_device(self.device) model = attempt_load(self.weights, device=self.device) model.to(self.device).eval() model.half() self.zwq = True self.model = model self.names = model.module.names if hasattr( model, 'module') else model.names def preprocess(self, img): img0 = img.copy() #img = cv2.resize(img, (1024, 768)) img = letterbox(img, new_shape=self.img_size, stride=self.stride)[0] img = img[:, :, ::-1].transpose(2, 0, 1) img = np.ascontiguousarray(img) img = torch.from_numpy(img).to(self.device) img = img.half() # ¡ã????¨¨ img /= 255.0 # ¨ª???1¨¦¨°??¡¥ if img.ndimension() == 3: img = img.unsqueeze(0) return img0, img def AIlabel(self,imagename): with torch.no_grad(): tqdm.write(imagename + " is doing!") image = cv2.imread(imagename) image_or = image img_detect = cv2.resize(image, (1024, 768)) # t1 = time.time() image_scale,image = self.preprocess(img_detect) # t2 = time.time() # if gp.DebugPrint_flag: # print("Detection data deal is ",(t2-t1)*1000) pred = self.model(image, augment=False)[0] pred = pred.float() pred = non_max_suppression(pred, self.conf_thres, self.iou_thres) # t3 = time.time() # if gp.DebugPrint_flag: # print("Detection inference is ",(t3-t2)*1000) boxes = [] scores = [] labels = [] for det in pred: if det is not None and len(det): det[:, :4] = scale_coords( image.shape[2:], det[:, :4], image_scale.shape).round() boxes.append(det[:, :4]) scores.append(det[:, 4]) labels.append((det[:, -1])) boxes = boxes[0] scores = scores[0] labels = labels[0] # t5 = time.time() # if gp.DebugPrint_flag: # print("Detection get result is ", (t5 - t3) * 1000) if len(labels) > 0: ratio_x = (image_or.shape[1] / 1024.0) ratio_y = (image_or.shape[0] / 768.0) for i in range(len(boxes)): boxes[i][0]=boxes[i][0] * ratio_x boxes[i][2]=boxes[i][2] * ratio_x boxes[i][1] = boxes[i][1] * ratio_y boxes[i][3] = boxes[i][3] * ratio_y if boxes[i][0]<0: boxes[i][0] = 0 if boxes[i][1]<0: boxes[i][1] = 0 if boxes[i][2]<0: boxes[i][2] = 0 if boxes[i][3]<0: boxes[i][3] = 0 # for i in range(len(labels)): # labels[i] = "bird" # t4 = time.time() # if gp.DebugPrint_flag: # print("Detection final is ", (t4 - t3) * 1000) xml_outputpath = imagename.replace("imagesAILabel","LabelXML").replace(".jpg",".xml") make_xml(imagename,image_or.shape[0],image_or.shape[1],boxes,labels,scores,xml_outputpath) shutil.copy(imagename, imagename.replace("imagesAILabel", "YesResult")) else: shutil.copy(imagename, imagename.replace("imagesAILabel", "NoResult")) def AIlabelFine(self, imagename, fine_grained): with torch.no_grad(): tqdm.write(imagename + " is doing!") try: image = cv2.imread(imagename) except: os.remove(imagename) return if not type(image): print("skip") return image_or = image try: img_detect = cv2.resize(image, (image.shape[1], image.shape[0])) except: shutil.copy(imagename, imagename.replace("imagesAILabel", "BrokenImage")) return image_scale, image = self.preprocess(img_detect) pred = self.model(image, augment=False)[0] pred = pred.float() pred = non_max_suppression(pred, self.conf_thres, self.iou_thres) boxes = [] scores = [] labels = [] for det in pred: if det is not None and len(det): det[:, :4] = scale_coords( image.shape[2:], det[:, :4], image_scale.shape).round() boxes.append(det[:, :4]) scores.append(det[:, 4]) labels.append((det[:, -1])) boxes = boxes[0] scores = scores[0] labels = labels[0] # if len(labels) > 0: ratio_x = 1#(image_or.shape[1] / 1024.0) ratio_y = 1#(image_or.shape[0] / 768.0) for i in range(len(boxes)): boxes[i][0] = boxes[i][0] * ratio_x boxes[i][2] = boxes[i][2] * ratio_x boxes[i][1] = boxes[i][1] * ratio_y boxes[i][3] = boxes[i][3] * ratio_y if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][3] < 0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 0 if self.zwq: boxes_fine, labels_fine, scores_fine = fine_grained.inference(image_or, boxes, labels, scores, 8) else: boxes_fine = boxes labels_fine = labels scores_fine = scores xml_outputpath = imagename.replace("imagesAILabel", "Annotations").replace(".jpg", ".xml") make_xml(imagename, image_or.shape[0], image_or.shape[1], boxes_fine, labels_fine, scores_fine, xml_outputpath) shutil.copy(imagename, imagename.replace("imagesAILabel", "JPEGImages")) # else: # shutil.copy(imagename, imagename.replace("imagesAILabel", "NoResult")) def inference(self, image, imagetime, index, fine_grained): with torch.no_grad(): rootpath = os.path.join(self.root, imagetime) if not os.path.exists(rootpath): os.mkdir(rootpath) if not os.path.exists(rootpath + "/" + "Detection" + "/"): os.mkdir(rootpath + "/" + "Detection" + "/") image_or = image img_detect = cv2.resize(image, (1024, 768)) image_scale, image = self.preprocess(img_detect) pred = self.model(image, augment=False)[0] pred = pred.float() pred = non_max_suppression(pred, self.conf_thres, self.iou_thres) boxes = [] scores = [] labels = [] for det in pred: if det is not None and len(det): det[:, :4] = scale_coords( image.shape[2:], det[:, :4], image_scale.shape).round() boxes.append(det[:, :4]) scores.append(det[:, 4]) labels.append((det[:, -1])) boxes = boxes[0] scores = scores[0] labels = labels[0] if len(labels) > 0: indices_label_left = labels < 10 indices_label_right = labels < 10 indices = [] for i in range(len(indices_label_left)): indices.append(indices_label_left[i] and indices_label_right[i]) boxes = boxes[indices] labels = labels[indices] scores = scores[indices] # ÒòΪĪÃûÆæÃî¶àÁËÒ»¸ölistά¶È£¬ËùÒÔ¼ÓÁËÕâ¸ö0 if len(labels) > 0: boxes = boxes[0] labels = labels[0] scores = scores[0] ratio_x = (3840.0 / 1024.0) ratio_y = (2160.0 / 768.0) for i in range(len(boxes)): boxes[i][0] = boxes[i][0] * ratio_x boxes[i][2] = boxes[i][2] * ratio_x boxes[i][1] = boxes[i][1] * ratio_y boxes[i][3] = boxes[i][3] * ratio_y if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][3] < 0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 99999 boxes_fine, labels_fine, scores_fine = fine_grained.inference(image_or, boxes, labels, scores) else: boxes_fine, labels_fine, scores_fine = [],[],[] if gp.DebugPrint_flag: print("Detection is nothong") xml_outputpath = rootpath + "/" + "Detection"+"/" + index+".xml" make_xml(str(index),image_or.shape[0],image_or.shape[1],boxes_fine,labels_fine,scores_fine,xml_outputpath) def inference_Red(self,image,ImagePath,fine_grained): with torch.no_grad(): image_or = image height, width = image.shape[:2] img_detect = cv2.resize(image, (1024, 768)) # t1 = time.time() image_scale, image = self.preprocess(img_detect) # t2 = time.time() # if gp.DebugPrint_flag: # print("Detection data deal is ", (t2 - t1) * 1000) pred = self.model(image, augment=False)[0] pred = pred.float() pred = non_max_suppression(pred, self.conf_thres, self.iou_thres) # t3 = time.time() # if gp.DebugPrint_flag: # print("Detection inference is ", (t3 - t2) * 1000) boxes = [] scores = [] labels = [] for det in pred: if det is not None and len(det): det[:, :4] = scale_coords( image.shape[2:], det[:, :4], image_scale.shape).round() boxes.append(det[:, :4]) scores.append(det[:, 4]) labels.append((det[:, -1])) boxes = boxes[0] scores = scores[0] labels = labels[0] # t5 = time.time() # if gp.DebugPrint_flag: # print("Detection get result is ", (t5 - t3) * 1000) # print(labels) if len(labels) > 0 : indices_label_left = labels < 1 indices_label_right = labels < 1 indices = [] for i in range(len(indices_label_left)): indices.append(indices_label_left[i] and indices_label_right[i]) boxes = boxes[indices] labels = labels[indices] scores = scores[indices] # ÒòΪĪÃûÆæÃî¶àÁËÒ»¸ölistά¶È£¬ËùÒÔ¼ÓÁËÕâ¸ö0 if len(labels) > 0: boxes = boxes[0] labels = labels[0] scores = scores[0] ratio_x = (width / 1024.0) ratio_y = (height / 768.0) for i in range(len(boxes)): boxes[i][0] = boxes[i][0] * ratio_x boxes[i][2] = boxes[i][2] * ratio_x boxes[i][1] = boxes[i][1] * ratio_y boxes[i][3] = boxes[i][3] * ratio_y if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][0] > 3840: boxes[i][0] = 3839 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][1] > 2160: boxes[i][1] = 2159 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][2] > 3840: boxes[i][2] = 3839 if boxes[i][3] < 0: boxes[i][3] = 0 if boxes[i][3] > 2160: boxes[i][3] = 2159 for i in range(len(labels)): labels[i] = 99999 boxes_fine, labels_fine, scores_fine = fine_grained.inference(image_or, boxes, labels, scores) # t4 = time.time() # if gp.DebugPrint_flag: # print("Detection final is ", (t4 - t3) * 1000) # t6 = time.time() xml_outputpath = ImagePath.replace("/ftp/", "/ftp_Result/").replace(".JPG", ".xml").replace(".jpg", ".xml") make_xml(str(ImagePath), image_or.shape[0], image_or.shape[1], boxes_fine, labels_fine, scores_fine, xml_outputpath) # Image.fromarray(drawn_image).save(xml_outputpath.replace(".xml",".jpg")) # t5 = time.time() # if gp.DebugPrint_flag: # print("Detection save is ", (t5 - t6) * 1000) else: boxes_fine, labels_fine, scores_fine = [], [], [] if gp.DebugPrint_flag: print("Detection is nothong") def check_object(self,img_o): flag = False with torch.no_grad(): # path = '/data/Hatcher/ouput1/0001_20210414151147/RawImage/0099.jpg' # img = cv2.imread(path) # img = cv2.resize() height, width = img_o.shape[:2] img = cv2.resize(img_o,(1024,768)) # t1 = time.time() image_scale, image = self.preprocess(img) # t2 = time.time() # if gp.DebugPrint_flag: # print("Check data deal is ", (t2 - t1) * 1000) pred = self.model(image, augment=False)[0] pred = pred.float() pred = non_max_suppression(pred, self.conf_thres, self.iou_thres) # t3 = time.time() # if gp.DebugPrint_flag: # print("Check inference is ", (t3 - t2) * 1000) boxes = [] scores = [] labels = [] for det in pred: if det is not None and len(det): det[:, :4] = scale_coords( image.shape[2:], det[:, :4], image_scale.shape).round() boxes.append(det[:, :4]) scores.append(det[:, 4]) labels.append((det[:, -1])) boxes = boxes[0] scores = scores[0] labels = labels[0] # t5 = time.time() # if gp.DebugPrint_flag: # print("Check get result is ", (t5 - t3) * 1000) if len(labels) > 0: # indices_label_left = labels < 1 # indices_label_right = labels < 1 # indices = [] # for i in range(len(indices_label_left)): # indices.append(indices_label_left[i] and indices_label_right[i]) # boxes = boxes[indices] # print(type(labels)) # print(labels) # print(type(indices)) # print(indices) labels = labels scores = scores # ÒòΪĪÃûÆæÃî¶àÁËÒ»¸ölistά¶È£¬ËùÒÔ¼ÓÁËÕâ¸ö0 if len(labels) > 0: if len(boxes.shape) == 3: boxes = boxes[0] labels = labels[0] scores = scores[0] ratio_x = (width / 1024.0) ratio_y = (height / 768.0) for i in range(len(boxes)): #boxes = boxes.numpy() boxes[i][0] = boxes[i][0] * ratio_x boxes[i][2] = boxes[i][2] * ratio_x boxes[i][1] = boxes[i][1] * ratio_y boxes[i][3] = boxes[i][3] * ratio_y if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][0] > 3840: boxes[i][0] = 3839 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][1] > 2160: boxes[i][1] = 2159 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][2] > 3840: boxes[i][2] = 3839 if boxes[i][3] < 0: boxes[i][3] = 0 if boxes[i][3] > 2160: boxes[i][3] = 2159 else: pass if 0 in labels: flag = True if gp.DebugPrint_flag: print("boxes = ", boxes) print("check flag is ", flag) return flag, labels, boxes, scores """ ϸÁ£¶Èʶ±ð """ class args_temp(): dataset = 'HCData' backbone = 'resnet50' batch_size = 1 num_workers = 8 version = 'val' resize_resolution = 512 crop_resolution = 448 save_suffix = None swap_num = [7, 7] class Fine_grained_T: def __init__(self, device): self.device = device self.WeightPath = gp.FineWeight self.net = timm.create_model("convnext_base_384_in22ft1k", num_classes=gp.FineClassNum, pretrained=False).to(device) # Multi_transform_image().to(device) model_dict = self.net.state_dict() pretrained_dict = torch.load(self.WeightPath) pretrained_dict = {k.replace("module.", ""): v for k, v in pretrained_dict.items()} model_dict.update(pretrained_dict) self.net.load_state_dict(pretrained_dict) self.transformers = load_data_transformers(600, 384, [7, 7]) print("load model weight is ok") self.net.train(False) self.softmax = nn.Softmax() self.gp_Species = torch.zeros(8).cuda(device) if gp.DebugPrint_flag: print("load model weight is ok") def inference(self, img, imagelist, labellist, scorelist, multiimagenum): with torch.no_grad(): img_deal = [] labelIndexForFusionResult = 0 imaIndexIfSuit = [0 for i in range(len(imagelist))] for i, box in enumerate(imagelist): if box[0] < 0: box[0] = 0 if box[1] < 0: box[1] = 0 if box[2] < 0: box[2] = 0 if box[3] < 0: box[3] = 0 Img = img.copy()[int(box[1]):int(box[3]), int(box[0]):int(box[2]), :] #if gp.DebugPrint_flag: #print(box) #print(Img) #if (int(box[3])-int(box[1])>gp.fine_thresh) and (int(box[2])-int(box[0])>gp.fine_thresh): if ((int(box[3]) - int(box[1])) * (int(box[2]) - int(box[0]))) > gp.fine_thresh1 : Img = Image.fromarray(cv2.cvtColor(Img, cv2.COLOR_BGR2RGB)) Img = self.transformers['test_totensor'](Img) Img = Img.unsqueeze(0) img_deal.append(Img) imaIndexIfSuit[i] =1 else: pass iter_nums = int(len(img_deal)/multiimagenum) single_imagedealnums = int(len(img_deal)%multiimagenum) for iter_num in range(iter_nums): tqdm.write("start {} iter".format(str(iter_num))) # t1 = time.time() Img = torch.cat((img_deal[0*iter_num], img_deal[1*iter_num], img_deal[2*iter_num], img_deal[3*iter_num], img_deal[4*iter_num], img_deal[5*iter_num], img_deal[6*iter_num], img_deal[7*iter_num]), 0) inputs = Variable(Img.cuda()) # t2 = time.time() # if gp.DebugPrint_flag: # print("iter{} fine data deal is ".format(str(iter_num)), (t2 - t1) * 1000) outputs = self.net(inputs) outputs = self.softmax(outputs) # t3 = time.time() # if gp.DebugPrint_flag: # print("iter{} fine inference is ".format(str(iter_num)), (t3 - t2) * 1000) #outputs[0] = self.softmax(outputs[0]) #Species = torch.zeros(multiimagenum).cuda() Species = self.gp_Species # t4_1 = time.time() # if gp.DebugPrint_flag: # print("iter{} fine data final1 is ".format(str(iter_num)), (t4_1 - t3) * 1000) for outnum in range(multiimagenum): top3_val, top3_pos = torch.topk(outputs[outnum],1) # t4_temp2 = time.time() if top3_val < gp.Fine_thresh: Species[outnum] = 0 else: Species[outnum] = top3_pos # t4_temp3 = time.time() # tqdm.write("iter{} fine data temp3 is ".format(str(iter_num)), (t4_temp3 - t4_temp2) * 1000) # t4_2 = time.time() # if gp.DebugPrint_flag: # tqdm.write("iter{} fine data final2 is ".format(str(iter_num)), (t4_2 - t4_1) * 1000) for j in range(multiimagenum): for m in range(len(imaIndexIfSuit)): if imaIndexIfSuit[labelIndexForFusionResult] == 1: #clam this boxes has been deal with fine grained labellist[labelIndexForFusionResult] = Species[j] labelIndexForFusionResult += 1 break else: labelIndexForFusionResult += 1 # t4_3 = time.time() # if gp.DebugPrint_flag: # tqdm.write("iter{} fine data final3 is ".format(str(iter_num)), (t4_3 - t4_2) * 1000) #labellist[iter_num*multiimagenum:iter_num*multiimagenum+multiimagenum] = Species # t4 = time.time() # if gp.DebugPrint_flag: # tqdm.write("iter{} fine data final is ".format(str(iter_num)), (t4 - t3) * 1000) # t4 = time.time() # if gp.DebugPrint_flag: # tqdm.write("start {} single_imagedealnums".format(str(single_imagedealnums))) # tqdm.write("single_imagedealnums is ",single_imagedealnums) Img = None if single_imagedealnums == 0: pass elif single_imagedealnums == 1: Img = img_deal[(iter_nums) * multiimagenum] else: #Img = torch.cat((img_deal[0*iter_num],img_deal[1*iter_num],img_deal[2*iter_num],img_deal[3*iter_num],img_deal[4*iter_num],img_deal[5*iter_num],img_deal[6*iter_num],img_deal[7*iter_num]),0) # tF1 = time.time() singleimage = [] for tempm in range(single_imagedealnums): singleimage.append(img_deal[(iter_nums) * multiimagenum+tempm]) Img = torch.cat(singleimage, 0) # Img = torch.cat((img_deal[(iter_nums) * multiimagenum], img_deal[(iter_nums) * multiimagenum+1]), 0) # for single_imagedealnum in range(single_imagedealnums-2): # Img = torch.cat((Img,img_deal[(iter_nums) * multiimagenum+1+single_imagedealnum+1]),0) # tF2 = time.time() # if gp.DebugPrint_flag: # tqdm.write("single_imagedealnumsv fine data deal cat is ", (tF2 - tF1) * 1000) # t5 = time.time() # if gp.DebugPrint_flag: # tqdm.write("single_imagedealnumsv fine data deal is ", (t5 - t4) * 1000) if not (Img == None): # tsfCU1 = time.time() inputs = Variable(Img.cuda()) # tsfCU2 = time.time() # if gp.DebugPrint_flag: # tqdm.write("single_imagedealnumsv tocuda is ", (tsfCU2 - tsfCU1) * 1000) outputs = self.net(inputs) outputs = self.softmax(outputs) # t6 = time.time() # if gp.DebugPrint_flag: # tqdm.write("single_imagedealnums fine inference is ", (t6 - t5) * 1000) Species = self.gp_Species #Species = torch.zeros(single_imagedealnums).cuda() for outnum in range(single_imagedealnums): top3_val, top3_pos = torch.topk(outputs[outnum], 1) #print("top3_val==========", top3_val) t4_temp2 = time.time() if top3_val < gp.Fine_thresh: Species[outnum] = 0 else: Species[outnum] = top3_pos # t4_temp3 = time.time() # print("single_imagedealnums fine data temp3 is ", (t4_temp3 - t4_temp2) * 1000) for j in range(single_imagedealnums): for m in range(len(imaIndexIfSuit)): if imaIndexIfSuit[labelIndexForFusionResult] == 1: # clam this boxes has been deal with fine grained labellist[labelIndexForFusionResult] = Species[j] labelIndexForFusionResult += 1 break else: labelIndexForFusionResult += 1 #labellist[len(labellist)-single_imagedealnums:] = Species # t7 = time.time() # if gp.DebugPrint_flag: # tqdm.write(" fine data final is ", (t7 - t6) * 1000) return imagelist, labellist, scorelist class Fine_grained: def __init__(self, device, root): self.device = device self.root = root self.DclPreWeightPath = '../data_MPKFNet/test/3.pth' # parser = argparse.ArgumentParser(description='dcl parameters') # parser.add_argument('--data', dest='dataset', default='HCData', type=str) # parser.add_argument('--backbone', dest='backbone', default='resnet50', type=str) # parser.add_argument('--b', dest='batch_size', default=1, type=int) # parser.add_argument('--nw', dest='num_workers', default=8, type=int) # parser.add_argument('--ver', dest='version', default='val', type=str) # parser.add_argument('--size', dest='resize_resolution', default=512, type=int) # parser.add_argument('--crop', dest='crop_resolution', default=448, type=int) # parser.add_argument('--ss', dest='save_suffix', default=None, type=str) # parser.add_argument('--acc_report', dest='acc_report', action='store_true') # parser.add_argument('--swap_num', default=[7, 7], nargs=2, metavar=('swap1', 'swap2'), type=int, help='specify a range') self.args = args_temp() self.Config = LoadConfig(self.args, 's') self.transformers = load_data_transformers(self.args.resize_resolution, self.args.crop_resolution, self.args.swap_num) self.net = MainModel(self.Config).to(device) model_dict = self.net.state_dict() pretrained_dict = torch.load(self.DclPreWeightPath) model_dict.update(pretrained_dict) self.net.load_state_dict(model_dict) if gp.DebugPrint_flag: print("load model weight is ok") self.net.train(False) def inference(self,img,imagelist,labellist,scorelist): with torch.no_grad(): for i, box in enumerate(imagelist): Img = img.copy()[int(box[1]):int(box[3]), int(box[0]):int(box[2]), :] if (int(box[3])-int(box[1]) > 100) and (int(box[2])-int(box[0]) > 100): t1 = time.time() Img = Image.fromarray(cv2.cvtColor(Img, cv2.COLOR_BGR2RGB)) Img = self.transformers['test_totensor'](Img) Img = Img.unsqueeze(0) inputs = Variable(Img.cuda()) # t2 = time.time() # if gp.DebugPrint_flag: # print("fine data deal is ", (t2 - t1) * 1000) outputs = self.net(inputs) # t3 = time.time() # if gp.DebugPrint_flag: # print("fine inference is ", (t3 - t2) * 1000) top3_val, top3_pos = torch.topk(outputs[0], 3) # topk·µ»ØµÄÊÇÒ»¸öÔª×飬µÚÒ»¸öÔªËØÊÇÿһÐÐ×î´óµÄÈý¸öÖµ£¨Ä¬ÈÏÊÇ°´ÐÐÈ¡×î´óÖµ£©£¬µÚ¶þ¸öÔªËØÊÇÿһÐÐ×î´óµÄÖµµÄÅÅÐò Species = top3_pos.cpu().numpy()[0][0] ConfidenceLevel = top3_val.cpu().numpy()[0][0] ConfidenceLevel = ConfidenceLevel/26.0 if ConfidenceLevel > 1: ConfidenceLevel = 1.0 * random.uniform(0, 1) # if gp.DebugPrint_flag: # print("hhhhhhhhhhhhhhhhhhhhhhh is ",ConfidenceLevel) labellist[i] = Species scorelist[i] = ConfidenceLevel # t4 = time.time() # if gp.DebugPrint_flag: # print("fine data final is ", (t4 - t3) * 1000) return imagelist, labellist, scorelist class Check_start (threading.Thread): def __init__(self, threadID, name, img, DetectionClass): threading.Thread.__init__(self) self.threadID = threadID self.name = name self.img = img self.DetectionClass = DetectionClass def run(self): if gp.DebugPrint_flag: tqdm.write(self.name + " is start") gp.is_ex_bird, labels, boxes, scores = self.DetectionClass.check_object(self.img) if gp.is_ex_bird: if gp.DebugPrint_flag: tqdm.write(gp.is_ex_bird) gp.image_time = gp.Camera_device +time.strftime('%Y%m%d%H%M%S', time.localtime(time.time())) if gp.DebugPrint_flag: tqdm.write(self.name + " is finish") class save_sigle_image(threading.Thread): def __init__(self, img): threading.Thread.__init__(self) self.img = img def run(self): cv2.imwrite( "../data_MPKFNet/RTSP2/" + "{}.jpg".format(time.strftime('%Y%m%d%H%M%S', time.localtime(time.time()))), self.img) class Check_start_fordataselect (threading.Thread): def __init__(self, threadID, name, img, DetectionClass, fine_grained): threading.Thread.__init__(self) self.threadID = threadID self.name = name self.img = img self.DetectionClass = DetectionClass self.fine_grained = fine_grained def run(self): if gp.DebugPrint_flag: print(self.name + " is start") is_ex_bird, labels, boxes, scores = self.DetectionClass.check_object(self.img) image_or = self.img.copy() if is_ex_bird: for i in range(len(boxes)): if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][3] < 0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 99999 boxes_fine, labels_fine, scores_fine = self.fine_grained.inference(self.img, boxes, labels, scores) for i in range(len(scores)): xyxy = boxes_fine[i] conf = scores_fine[i] cls = labels_fine[i] print(int(cls)) label = str(gp.BHG_chineseName[int(cls)]) print(label) #label = f'{int(cls)} {conf:.2f}' palette = (2 ** 11 - 1, 2 ** 15 - 1, 2 ** 20 - 1) colors = [int((p * (cls ** 2 - cls + 1)) % 255) for p in palette] self.img = plot_one_box(xyxy, self.img, label=label, color=colors, line_thickness=2) #cv2.imshow("AI-ECO",self.img) #cv2.waitKey(0) cv2.imwrite("../data_MPKFNet/RTSP/" + "{}.jpg".format(time.strftime('%Y%m%d%H%M%S', time.localtime(time.time()))), image_or) cv2.imwrite("../data_MPKFNet/RTSP/result/" + "{}.jpg".format(time.strftime('%Y%m%d%H%M%S', time.localtime(time.time()))), self.img) if gp.DebugPrint_flag: print(self.name + " is finish") class shishi(): def __init__(self, name, DetectionClass, fine_grained): self.threadID = 0 self.name = name self.DetectionClass = DetectionClass self.fine_grained = fine_grained def go_go(self, img, num_image, num_detection): spiec_number = [0 for i in range(len(gp.BHG_chineseName))] img_show = img.copy() boxes_fine = [] labels_fine = [] if (isinstance(img_show, np.ndarray)): # ?D??¨º?¡¤?OpenCV¨ª???¨¤¨¤D¨ª img_show = Image.fromarray(cv2.cvtColor(img_show, cv2.COLOR_BGR2RGB)) draw = ImageDraw.Draw(img_show) height_or, width_or = img.shape[:2] if gp.DebugPrint_flag: print("image size is ", height_or, width_or) if(num_detection%2==0): if gp.DebugPrint_flag: print(self.name + " is start") # ts1 = time.time() is_ex_bird, labels, boxes, scores = self.DetectionClass.check_object(img) # ts2 = time.time() # if gp.DebugPrint_flag: # print("shishi time detect1 deal is ", (ts2 - ts1) * 1000) if is_ex_bird: for i in range(len(boxes)): if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][3] < 0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 0#gp.birdIndex # ts3 = time.time() # if gp.DebugPrint_flag: # print("shish time detect2 deal is ", (ts3 - ts2) * 1000) boxes_fine, labels_fine, scores_fine = self.fine_grained.inference(img, boxes, labels, scores, 8) # ts4 = time.time() # if gp.DebugPrint_flag: # print("shish time detect3 deal is ", (ts4 - ts3) * 1000) # tcpu1 = time.time() boxes_fine = boxes_fine.to('cpu') labels_fine = labels_fine.to('cpu') # tcpu2 = time.time() # if gp.DebugPrint_flag: # print("to cpu timeis ", (tcpu2 - tcpu1) * 1000) gp.boxes_fine = boxes_fine gp.labels_fine = labels_fine else: print("don't deal {}".format(str(num_detection))) boxes_fine = gp.boxes_fine labels_fine = gp.labels_fine for i in range(len(boxes_fine)): # tprint1 = time.time() xyxy = boxes_fine[i] cls = labels_fine[i] # tprint2 = time.time() # if gp.DebugPrint_flag: # print("tprint1 is ", (tprint2 - tprint1) * 1000) # tcls1 = time.time() b = (((int(xyxy[3]) - int(xyxy[1])) * (int(xyxy[2]) - int(xyxy[0]))) < 15000) # a =(xyxy.item()[3] - xyxy.item()[1]) * (xyxy.item()[2] - xyxy.item()[0]) # tcls2 = time.time() # if gp.DebugPrint_flag: # print("tcls1 is ", (tcls2 - tcls1) * 1000) if cls == 0 or (((int(xyxy[3]) - int(xyxy[1])) * (int(xyxy[2]) - int(xyxy[0]))) < gp.fine_thresh2): #if cls == gp.birdIndex or (torch.sub(torch.mul(torch.sub(xyxy[3], xyxy[1]), torch.sub(xyxy[2], xyxy[0])),gp.fine_thresh2) < 0): label = None else: # tcls3 = time.time() label = gp.BHG_chineseName[int(cls)] spiec_number[int(cls)] += 1 # tcls4 = time.time() # if gp.DebugPrint_flag: # print("tcls2 is ", (tcls4 - tcls3) * 1000) # tprint3 = time.time() # if gp.DebugPrint_flag: # print("tprint2 is ", (tprint3 - tprint2) * 1000) #label = f'{int(cls)} {conf:.2f}' palette = (2 ** 11 - 1, 2 ** 15 - 1, 2 ** 20 - 1) colors = gp.color_box#tuple(int((p * (4 ** 2 - 4 + 1)) % 220) for p in palette) # tprint4 = time.time() # if gp.DebugPrint_flag: # print("tprint3 is ", (tprint4 - tprint3) * 1000) draw = plot_one_box(xyxy, draw, height_or,width_or,label=label, color=colors, line_thickness=2) # tpChin1 = time.time() if num_image > gp.showThresh : gp.num_count_old = len(boxes_fine) gp.spiec_number_old = spiec_number num_image = 0 # tpChin2 = time.time() # if gp.DebugPrint_flag: # print("tpChin1 is ", (tpChin2 - tpChin1) * 1000) #if width_or == 2560: draw = cv2AddChineseText(draw, "ʶ±ð×ÜÊý: {}Ö»".format(str(gp.num_count_old)), (50*gp.radioShowX, 1100*gp.radioShowY), textColor=gp.color_show, textSize=50) # tpChin3 = time.time() # if gp.DebugPrint_flag: # print("tpChin2 is ", (tpChin3 - tpChin2) * 1000) draw = cv2AddChineseText(draw, "·ÖÀàʶ±ð:", (50*gp.radioShowX, 1150*gp.radioShowY), textColor=gp.color_show, textSize=50) # tpChin4 = time.time() # if gp.DebugPrint_flag: # print("tpChin3 is ", (tpChin4 - tpChin3) * 1000) for i in range(1, len(spiec_number)): if i < 5: draw = cv2AddChineseText(draw, "{}:{}Ö»".format(str(gp.BHG_chineseName[i]), str(gp.spiec_number_old[i])), (300*gp.radioShowX + (i-1) * 280*gp.radioShowX, 1150*gp.radioShowY), textColor=gp.color_show, textSize=50) else: draw = cv2AddChineseText(draw, "{}:{}Ö»".format(str(gp.BHG_chineseName[i]), str(gp.spiec_number_old[i])), (50*gp.radioShowX + (i-5) * 280*gp.radioShowX, 1200*gp.radioShowY), textColor=gp.color_show, textSize=50) img_show = cv2.cvtColor(np.asarray(img_show), cv2.COLOR_RGB2BGR) gp.img_show = img_show # tpChin2 = time.time() # if gp.DebugPrint_flag: # print("tpChin is ", (tpChin2 - tpChin1) * 1000) #cv2.waitKey(0) # cv2.imwrite("../data_MPKFNet/RTSP/" + "{}.jpg".format(time.strftime('%Y%m%d%H%M%S', time.localtime(time.time()))), # image_or) # cv2.imwrite( # "../data_MPKFNet/RTSP/result/" + "{}.jpg".format(time.strftime('%Y%m%d%H%M%S', time.localtime(time.time()))), # self.img) if gp.DebugPrint_flag: print(self.name + " is finish") num_image += 1 print(num_image) return img_show, num_image class RealDeal(): def __init__(self, name, DetectionClass, fine_grained): self.threadID = 0 self.name = name self.DetectionClass = DetectionClass self.fine_grained = fine_grained def realtimerestart(self, img, num_image, num_detection, SortRestart): spiec_number = [0 for i in range(len(gp.BHG_chineseName))] img_show = img.copy() boxes_fine = [] labels_fine = [] if (isinstance(img_show, np.ndarray)): # ?D??¨º?¡¤?OpenCV¨ª???¨¤¨¤D¨ª img_show = Image.fromarray(cv2.cvtColor(img_show, cv2.COLOR_BGR2RGB)) draw = ImageDraw.Draw(img_show) height_or, width_or = img.shape[:2] # if gp.DebugPrint_flag: # print("image size is ",height_or,width_or) # if gp.DebugPrint_flag: # print(self.name + " is start") # ts1 = time.time() is_ex_bird, labels, boxes, scores = self.DetectionClass.check_object(img) # ts2 = time.time() # if gp.DebugPrint_flag: # print("shishi time detect1 deal is ", (ts2 - ts1) * 1000) if is_ex_bird: for i in range(len(boxes)): if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][3] < 0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 0#gp.birdIndex # ts3 = time.time() # if gp.DebugPrint_flag: # print("shish time detect2 deal is ", (ts3 - ts2) * 1000) boxes_fine, labels_fine, scores_fine = self.fine_grained.inference(img, boxes, labels, scores,8) # ts4 = time.time() # if gp.DebugPrint_flag: # print("shish time detect3 deal is ", (ts4 - ts3) * 1000) # tcpu1 = time.time() boxes_fine = boxes_fine.to('cpu') labels_fine = labels_fine.to('cpu') # tcpu2 = time.time() # if gp.DebugPrint_flag: # print("to cpu timeis ", (tcpu2 - tcpu1) * 1000) gp.boxes_fine = boxes_fine gp.labels_fine = labels_fine #input = [x1,y1,x2,y2,kong,label] temp_sort = [i + 100 for i in range(len(labels_fine))] DetFine = np.column_stack((boxes_fine, temp_sort, labels_fine)) boxes_fine, indexmaxtemp, newObjectNum = SortRestart.update(DetFine) # boxes_fine ÊÇ n*6µÄÊý×飬nÊÇÄ¿±êµÄÊýÁ¿£¬¶øÇÒÐòºÅÊǵ¹ÐòµÄ£¬×îСµÄÐòºÅÊÇ1 if indexmaxtemp > gp.IndexMax: gp.IndexMax = indexmaxtemp #boxes_fine = boxes_fine[0] #output = [x1,y1,x2,y2,label,index] gp.Finerestartboxes.append(boxes_fine) # for i in range(len(boxes_fine)): # xyxy = boxes_fine[i][:4] # cls = boxes_fine[i][4] # index = boxes_fine[i][5] # label = None # palette = (2 ** 11 - 1, 2 ** 15 - 1, 2 ** 20 - 1) # colors = gp.color_box #tuple(int((p * (4 ** 2 - 4 + 1)) % 220) for p in palette) # draw = plot_one_box(xyxy, draw, height_or,width_or,label=label, Index= index, color=colors, line_thickness=int(3*gp.radioShowtext)) # if num_image > gp.showThresh : # gp.num_count_old = len(boxes_fine) # gp.spiec_number_old = spiec_number # num_image = 0 # #if width_or == 2560:识别总数分类识别 # draw = cv2AddChineseText(draw, "识别总数: {}只".format(str(gp.num_count_old)), # (50 * gp.radioShowX, 1100 * gp.radioShowY), # textColor=gp.color_show, # textSize=int(50* gp.radioShowtext)) # tpChin3 = time.time() # # draw = cv2AddChineseText(draw, "分类识别:", (50 * gp.radioShowX, 1150 * gp.radioShowY), textColor=gp.color_show, # textSize=int(50* gp.radioShowtext)) # tpChin4 = time.time() # if gp.DebugPrint_flag: # print("tpChin3 is ", (tpChin4 - tpChin3) * 1000) # for i in range(1, len(spiec_number)): # if i < 5: # draw = cv2AddChineseText(draw, "{}:{}只»".format(str(gp.BHG_chineseName[i]), str(gp.spiec_number_old[i])), # (300 * gp.radioShowX + (i - 1) * 280 * gp.radioShowX, 1150 * gp.radioShowY), # textColor=gp.color_show, # textSize=int(50* gp.radioShowtext)) # else: # draw = cv2AddChineseText(draw, "{}:{}只".format(str(gp.BHG_chineseName[i]), str(gp.spiec_number_old[i])), # (50 * gp.radioShowX + (i - 5) * 280 * gp.radioShowX, 1200 * gp.radioShowY), # textColor=gp.color_show, # textSize=int(50* gp.radioShowtext)) img_show = cv2.cvtColor(np.asarray(img_show), cv2.COLOR_RGB2BGR) gp.img_show = img_show tpChin2 = time.time() if gp.DebugPrint_flag: print(self.name + " is finish") num_image += 1 return img_show, num_image def realtime(self, img, num_image, num_detection, SortRestart): spiec_number = [0 for i in range(len(gp.BHG_chineseName))] img_show = img.copy() boxes_fine = [] labels_fine = [] if (isinstance(img_show, np.ndarray)): # ?D??¨º?¡¤?OpenCV¨ª???¨¤¨¤D¨ª img_show = Image.fromarray(cv2.cvtColor(img_show, cv2.COLOR_BGR2RGB)) draw = ImageDraw.Draw(img_show) height_or, width_or = img.shape[:2] if gp.DebugPrint_flag: print("image size is ", height_or, width_or) if gp.DebugPrint_flag: print(self.name + " is start") # ts1 = time.time() is_ex_bird, labels, boxes, scores = self.DetectionClass.check_object(img) # ts2 = time.time() # if gp.DebugPrint_flag: # print("shishi time detect1 deal is ", (ts2 - ts1) * 1000) if is_ex_bird : # print("Ä¿±ê¼ì²âµÄ¸öÊý ", len(boxes)) for i in range(len(boxes)): if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][3] < 0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 0#gp.birdIndex # ts3 = time.time() # if gp.DebugPrint_flag: # print("shish time detect2 deal is ", (ts3 - ts2) * 1000) # ts4 = time.time() # if gp.DebugPrint_flag: # print("shish time detect3 deal is ", (ts4 - ts3) * 1000) # tcpu1 = time.time() boxes_fine = boxes.to('cpu') labels_fine = labels.to('cpu') # tcpu2 = time.time() # if gp.DebugPrint_flag: # print("to cpu timeis ", (tcpu2 - tcpu1) * 1000) #############################Ä¿±ê×·×Ù£¬½«×·×ÙºóµÄ½á¹û·ÖΪ»÷ÖкÍδ»÷ÖÐÁ½²¿·Ö£¬Ç°ÕßÖ±½Ó´ÓÈ«¾Ö±äÁ¿SortLabel_IndexÖлñÈ¡label£¬ºóÕßÐèÒªÖØнøÐÐϸÁ£¶Èʶ±ð£¬×îºó½«Á½Õß½øÐÐÆ´½Ó #Êä³öΪboxes_fine [x1,y1,x2,y2,label,Index] temp_sort = [i + 100 for i in range(len(labels_fine))] DetFine = np.column_stack((boxes_fine, temp_sort, labels_fine)) boxes_fine, indexmaxtemp, newObjectNum = SortRestart.update(DetFine) IndexList = boxes_fine[:, -1].tolist() #½«indexÁÐÈ¡³öÀ´£¬Åжϵ±Ç°Ö¡¼ì²âµ½Á˶àÉÙÄ¿±ê # newObjectNum = 0 # if indexmaxtemp > gp.IndexMax: # newObjectNum = indexmaxtemp - gp.IndexMax # gp.IndexMax = indexmaxtemp # print(indexmaxtemp,newObjectNum) if (len(gp.SortLabel_Index) == 0): if (len(boxes_fine)!=0): boxes_fine_unmatched = [] # ÓÃÀ´´æ·Åµ±Ç°Ö¡Ã»Óл÷ÖеÄÄ¿±ê [xyxy, label,Index] for c in IndexList: boxes_fine_unmatched.append(boxes_fine[IndexList.index(c),:]) # δƥÅäµÄÄ¿±ê£¬ÐèÒª½øÒ»²½½øÐÐϸÁ£¶Èʶ±ð,ÓÃindexË÷Òýµ½ÊÇÄĸöÄ¿±ê£¬²¢½«¸ÃÄ¿±êµÄÊý¾Ý¶¼appendµ½unmatchÖÐ # ¶Ô¼¯ÖÐÆ¥ÅäµÄÄ¿±ê£¬½«labelÌ滻ΪSortLabel_IndexÖеÄlabel£¬ÐèÒªÕë¶Ôindex½øÐÐË÷Òý boxes_fine_unmatched = np.array(boxes_fine_unmatched) # boxes_fine_matched ÓÃÓÚ×îºóÆ´½Ó [xyxy,label,index] #####½øÐÐϸÁ£¶Èʶ±ð boxes = boxes_fine_unmatched[:, 0:4] labels = boxes_fine_unmatched[:, 4] index_fine_unF = boxes_fine_unmatched[:, 5] boxes_fine_unF, labels_fine_unF, index_fine_unF = self.fine_grained.inference(img, boxes, labels, index_fine_unF, 8) # print(labels_fine_unF) # ¸ñʽ [[xyxy]] [label1,label2...] [index1,index2...] # ÏȶÔмì²âµÄÄ¿±ê½øÐмì²é£¬ÊÇ·ñÒѾ­´ïµ½5´ÎÒÔÉÏʶ±ð ###¶ÔϸÁ£¶Èʶ±ð½á¹û½øÐкó´¦Àí boxes_fine_unmatched[:, 4] = labels_fine_unF boxes_fine_unmatched_temp = [] for i in range(len(labels_fine_unF)): if (len(gp.SortLabel_Indexpre)==0 ): # ˵Ã÷preÖÐûÓÐÄ¿±ê£¬Ôò¿ÉÒÔÖ±½Ó½«µ±Ç°µÄÄ¿±ê¼Ó½øÀ´ temp_a = np.array([labels_fine_unF[i], index_fine_unF[i], 1]) gp.SortLabel_Indexpre = np.expand_dims(np.append(gp.SortLabel_Indexpre, temp_a, axis=0), axis=0) else: # ˵Ã÷±äÁ¿ÖÐÒѾ­ÓÐÄ¿±ê£¬ÐèÒª¶Ôÿ¸öÄ¿±êµÄ´ÎÊý½øÐмì²â temp_a = np.array([labels_fine_unF[i], index_fine_unF[i], 1]) temp_index = index_fine_unF[i] # µ±Ç°Ä¿±êµÄindex index_list = gp.SortLabel_Indexpre[:, 1].tolist() # ÓÃÀ´Çó½âindex=nÔÚÊý×éÖеÄά¶ÈÐÅÏ¢£¬µÚ¼¸ÐÐ if temp_index in gp.SortLabel_Indexpre[:, 1]: # ˵Ã÷δ¼ì²âµÄÄ¿±êÔÚpreÈ«¾ÖÖУ¬ÐèÒª½øÒ»²½¼ì²âÊÇ·ñ¼ì²â´ÎÊý³¬¹ýãÐÖµ´ÎÊý # print("ϸÁ£¶Èʶ±ð´ÎÊý", gp.SortLabel_Indexpre[:, -1][index_list.index(temp_index)], gp.FineCountThresh) if (gp.SortLabel_Indexpre[:, -1][index_list.index(temp_index)] >= gp.FineCountThresh): # ÔÚpreÈ«¾Ö±äÁ¿ÖУ¬µ±Ç°indexµÄÄ¿±ê£¬³öÏֵĴ˴¦ÒѾ­³¬¹ýãÐÖµ£¬ÔòÖ±½Ó½«¸ÃindexºÍ¶ÔÓ¦µÄlabel£¬appendµ½label_indexÈ«¾Ö±äÁ¿ÖУ¬²¢¸üе½µ±Ç°¼ì²â½á¹ûÖУ¬²¢½«¸ÃindexÔÚpreÖÐɾ³ýµô index_dontthresh = [] if (len(gp.SortLabel_Index.shape) == 1): # мì²â³öÒ»¸öÄ¿±êµÄʱºò£¬ÐèÒª¶ÔÈ«¾Ö±äÁ¿SortLabel_Index½øÐг¤¶È¼ì²â£¬Èç¹ûÔ­ÏÈÀïÃæûÓУ¬Ôò²»¶Ôa´¦Àí£¬ÓеĻ°£¬ÐèÒª¶Ôa½øÐÐÌí¼Óά¶È´¦Àí a = np.array([labels_fine_unF[i], index_fine_unF[i]]) #print("a ======", a) #print("gp.SortLabel_Index ===", gp.SortLabel_Index) # if a[1] in gp.SortLabel_Index[:,1]: #˵Ã÷indexÒѾ­ÔÚÀïÃæÁË gp.SortLabel_Index = np.expand_dims( np.append(gp.SortLabel_Index, a, axis=0), axis=0) #print("gp.SortLabel_Indexed ===", gp.SortLabel_Index) gp.SortLabel_Indexpre = np.delete(gp.SortLabel_Indexpre,[index_list.index(temp_index)], axis=0) # gp.SortLabel_Indexpre[index_list.index(temp_index),1] = 0 else: a = np.expand_dims(np.array([labels_fine_unF[i], index_fine_unF[i]]), axis=0) #print("a ======", a) #print("gp.SortLabel_Index ===", gp.SortLabel_Index) # if a[1] in gp.SortLabel_Index[:,1]: #˵Ã÷indexÒѾ­ÔÚÀïÃæÁË gp.SortLabel_Index = np.concatenate((gp.SortLabel_Index, a), axis=0) #print("gp.SortLabel_Indexed ===", gp.SortLabel_Index) gp.SortLabel_Indexpre = np.delete(gp.SortLabel_Indexpre, [index_list.index(temp_index)], axis=0) #gp.SortLabel_Indexpre[index_list.index(temp_index), 1] = 0 else: # ÔÚpreÈ«¾Ö±äÁ¿ÖУ¬µ±Ç°indexµÄÄ¿±ê£¬³öÏֵĴ˴¦Î´³¬¹ýãÐÖµ gp.SortLabel_Indexpre[:, -1][index_list.index(temp_index)] += 1 # boxes_fine_unmatched = np.delete(boxes_fine_unmatched, [boxes_fine_unmatched[:, -1].tolist().index(temp_index)], # axis=0) # ½«µ±Ç°Ä¿±êÐÅÏ¢´Óunmatched Êý×éÖÐɾ³ýµô boxes_fine_unmatched_temp = boxes_fine_unmatched boxes_fine_unmatched = np.delete(boxes_fine_unmatched, [ boxes_fine_unmatched[:, -1].tolist().index(temp_index)], axis=0) boxes_fine_unmatched_temp[ boxes_fine_unmatched_temp[:, -1].tolist().index(temp_index), 4] = 0 else: # ²»ÔÚpreÖУ¬Ö±½Ó½«µ±Ç°µÄÄ¿±êappendµ½preÈ«¾Ö±äÁ¿ÖÐ gp.SortLabel_Indexpre = np.concatenate((gp.SortLabel_Indexpre, np.expand_dims(temp_a, axis=0)), axis=0) # boxes_fine_unmatched = np.column_stack((boxes_fine_unF,labels_fine_unF)) #½«ÐÂʶ±ðµÄϸÁ£¶È½á¹ûºÍ±ß¿òÈںϵ½Ò»Æ𣬺ÍÔ­ÏȵĸñʽһÖ # boxes_fine_unmatched = np.column_stack((boxes_fine_unmatched,index_fine_unF)) boxes_fine = boxes_fine_unmatched gp.SortLabel_Index = boxes_fine[:, 4:] # ¸üÐÂÈ«¾Ö±äÁ¿ÖеÄÐÅÏ¢ #######ÐÞ¸Äbug£¬ÔÚ¾ö²ß¹ý³ÌÖУ¬Òª°Ñ±ß¿òÏÔʾ³öÀ´ if len(boxes_fine_unmatched_temp) != 0: boxes_fine = boxes_fine_unmatched_temp else: if [True for c in IndexList if c not in gp.SortLabel_Index[:, -1]]:#˵Ã÷ÓÐмì²â³öÀ´µÄ£¬ÐèÒª½øÐÐϸÁ£¶Èʶ±ð£¬²¢½«Ð´¦ÀíµÄ½á¹ûºÍindexÆ¥ÅäÒ»Æð¸üе½SortLabel_IndexÈ«¾Ö±äÁ¿ÀïÃæ boxes_fine_unmatched = [] #ÓÃÀ´´æ·Åµ±Ç°Ö¡Ã»Óл÷ÖеÄÄ¿±ê [xyxy, label,Index] boxes_fine_matched = [] #ÓÃÀ´´æ·Åµ±Ç°Ö¡»÷ÖеÄÄ¿±ê [xyxy,label,index] for c in IndexList: if c not in gp.SortLabel_Index[:, -1]: boxes_fine_unmatched.append(boxes_fine[IndexList.index(c), :]) #δƥÅäµÄÄ¿±ê£¬ÐèÒª½øÒ»²½½øÐÐϸÁ£¶Èʶ±ð,ÓÃindexË÷Òýµ½ÊÇÄĸöÄ¿±ê£¬²¢½«¸ÃÄ¿±êµÄÊý¾Ý¶¼appendµ½unmatchÖÐ else: boxes_fine_matched.append(boxes_fine[IndexList.index(c), :]) #»÷ÖÐÆ¥ÅäµÄÄ¿±ê #¶Ô¼¯ÖÐÆ¥ÅäµÄÄ¿±ê£¬½«labelÌ滻ΪSortLabel_IndexÖеÄlabel£¬ÐèÒªÕë¶Ôindex½øÐÐË÷Òý boxes_fine_unmatched = np.array(boxes_fine_unmatched) boxes_fine_matched = np.array(boxes_fine_matched) for match_i in range(boxes_fine_matched.shape[0]): SortLabel_Indextemp = gp.SortLabel_Index[:, 1].tolist() if gp.DebugPrint_flag: print(SortLabel_Indextemp) boxes_fine_matched[match_i][4] = gp.SortLabel_Index[SortLabel_Indextemp.index(boxes_fine_matched[match_i][5]), 0] #boxes_fine_matched ÓÃÓÚ×îºóÆ´½Ó [xyxy,label,index] #####½øÐÐϸÁ£¶Èʶ±ð boxes = boxes_fine_unmatched[:, 0:4] labels = boxes_fine_unmatched[:, 4] index_fine_unF = boxes_fine_unmatched[:, 5] boxes_fine_unF, labels_fine_unF, index_fine_unF = self.fine_grained.inference(img, boxes, labels, index_fine_unF, 8) #¸ñʽ [[xyxy]] [label1,label2...] [index1,index2...] #ÏȶÔмì²âµÄÄ¿±ê½øÐмì²é£¬ÊÇ·ñÒѾ­´ïµ½5´ÎÒÔÉÏʶ±ð ###¶ÔϸÁ£¶Èʶ±ð½á¹û½øÐкó´¦Àí boxes_fine_unmatched[:, 4] = labels_fine_unF boxes_fine_unmatched_temp = [] for i in range(len(labels_fine_unF)): if (len(gp.SortLabel_Indexpre) == 0): #˵Ã÷preÖÐûÓÐÄ¿±ê£¬Ôò¿ÉÒÔÖ±½Ó½«µ±Ç°µÄÄ¿±ê¼Ó½øÀ´ temp_a = np.array([labels_fine_unF[i], index_fine_unF[i], 1]) gp.SortLabel_Indexpre = np.expand_dims(np.append(gp.SortLabel_Indexpre, temp_a, axis=0), axis=0) else: #˵Ã÷±äÁ¿ÖÐÒѾ­ÓÐÄ¿±ê£¬ÐèÒª¶Ôÿ¸öÄ¿±êµÄ´ÎÊý½øÐмì²â temp_a = np.array([labels_fine_unF[i], index_fine_unF[i], 1]) temp_index = index_fine_unF[i] #µ±Ç°Ä¿±êµÄindex index_list = gp.SortLabel_Indexpre[:, 1].tolist() #ÓÃÀ´Çó½âindex=nÔÚÊý×éÖеÄά¶ÈÐÅÏ¢£¬µÚ¼¸ÐÐ if temp_index in gp.SortLabel_Indexpre[:, 1]: #˵Ã÷δ¼ì²âµÄÄ¿±êÔÚpreÈ«¾ÖÖУ¬ÐèÒª½øÒ»²½¼ì²âÊÇ·ñ¼ì²â´ÎÊý³¬¹ýãÐÖµ´ÎÊý print("ϸÁ£¶Èʶ±ð´ÎÊý", gp.SortLabel_Indexpre[:, -1][index_list.index(temp_index)], gp.FineCountThresh) if (gp.SortLabel_Indexpre[:, -1][index_list.index(temp_index)] >= gp.FineCountThresh ): #ÔÚpreÈ«¾Ö±äÁ¿ÖУ¬µ±Ç°indexµÄÄ¿±ê£¬³öÏֵĴ˴¦ÒѾ­³¬¹ýãÐÖµ£¬ÔòÖ±½Ó½«¸ÃindexºÍ¶ÔÓ¦µÄlabel£¬appendµ½label_indexÈ«¾Ö±äÁ¿ÖУ¬²¢¸üе½µ±Ç°¼ì²â½á¹ûÖУ¬²¢½«¸ÃindexÔÚpreÖÐɾ³ýµô if (len(gp.SortLabel_Index) == 0): # мì²â³öÒ»¸öÄ¿±êµÄʱºò£¬ÐèÒª¶ÔÈ«¾Ö±äÁ¿SortLabel_Index½øÐг¤¶È¼ì²â£¬Èç¹ûÔ­ÏÈÀïÃæûÓУ¬Ôò²»¶Ôa´¦Àí£¬ÓеĻ°£¬ÐèÒª¶Ôa½øÐÐÌí¼Óά¶È´¦Àí a = np.array([labels_fine_unF[i], index_fine_unF[i]]) print("a ======", a) print("gp.SortLabel_Index ===", gp.SortLabel_Index) # if a[1] in gp.SortLabel_Index[:,1]: #˵Ã÷indexÒѾ­ÔÚÀïÃæÁË gp.SortLabel_Index = np.expand_dims(np.concatenate((gp.SortLabel_Index, a), axis=0), axis=0) print("gp.SortLabel_Indexed ===", gp.SortLabel_Index) gp.SortLabel_Indexpre = np.delete(gp.SortLabel_Indexpre, [index_list.index(temp_index)], axis=0) else: a = np.expand_dims(np.array([labels_fine_unF[i], index_fine_unF[i]]), axis=0) print("a ======", a) print("gp.SortLabel_Index ===", gp.SortLabel_Index) # if a[1] in gp.SortLabel_Index[:,1]: #˵Ã÷indexÒѾ­ÔÚÀïÃæÁË gp.SortLabel_Index = np.concatenate((gp.SortLabel_Index, a), axis=0) print("gp.SortLabel_Indexed ===", gp.SortLabel_Index) gp.SortLabel_Indexpre = np.delete(gp.SortLabel_Indexpre, [index_list.index(temp_index)], axis=0) else: boxes_fine_unmatched_temp = boxes_fine_unmatched gp.SortLabel_Indexpre[:, -1][index_list.index(temp_index)] += 1 boxes_fine_unmatched = np.delete(boxes_fine_unmatched, [boxes_fine_unmatched[:, -1].tolist().index(temp_index)], axis=0) #½«µ±Ç°Ä¿±êÐÅÏ¢´Óunmatched Êý×éÖÐɾ³ýµô boxes_fine_unmatched_temp[boxes_fine_unmatched_temp[:, -1].tolist().index(temp_index), 4] = 0 else: #²»ÔÚpreÖУ¬Ö±½Ó½«µ±Ç°µÄÄ¿±êappendµ½preÈ«¾Ö±äÁ¿ÖÐ gp.SortLabel_Indexpre = np.concatenate((gp.SortLabel_Indexpre, np.expand_dims(temp_a, axis=0)), axis=0) boxes_fine_unmatched = np.delete(boxes_fine_unmatched, [boxes_fine_unmatched[:, -1].tolist().index(temp_index)], axis=0) # ½«µ±Ç°Ä¿±êÐÅÏ¢´Óunmatched Êý×éÖÐɾ³ýµô # boxes_fine_unmatched = np.column_stack((boxes_fine_unF,labels_fine_unF)) #½«ÐÂʶ±ðµÄϸÁ£¶È½á¹ûºÍ±ß¿òÈںϵ½Ò»Æ𣬺ÍÔ­ÏȵĸñʽһÖ # boxes_fine_unmatched = np.column_stack((boxes_fine_unmatched,index_fine_unF)) if (len(boxes_fine_matched) == 0): boxes_fine = boxes_fine_unmatched elif (len(boxes_fine_unmatched) == 0): boxes_fine = boxes_fine_matched else: boxes_fine = np.concatenate((boxes_fine_matched, boxes_fine_unmatched)) gp.SortLabel_Index = boxes_fine[:, 4:] #¸üÐÂÈ«¾Ö±äÁ¿ÖеÄÐÅÏ¢ #######ÐÞ¸Äbug£¬ÔÚ¾ö²ß¹ý³ÌÖУ¬Òª°Ñ±ß¿òÏÔʾ³öÀ´ if len(boxes_fine_unmatched_temp) != 0: if (len(boxes_fine_matched) == 0): boxes_fine = boxes_fine_unmatched_temp else: boxes_fine = np.concatenate((boxes_fine_matched, boxes_fine_unmatched_temp)) else: for match_i in range(boxes_fine.shape[0]): if gp.DebugPrint_flag: print("gp.SortLabel_Index[:,1].tolist() = ", gp.SortLabel_Index[:, 1].tolist()) print("boxes_fine[match_i][5] = ", boxes_fine[match_i][5]) SortLabel_Indextemp = gp.SortLabel_Index[:, 1].tolist() boxes_fine[match_i][4] = gp.SortLabel_Index[SortLabel_Indextemp.index(boxes_fine[match_i][5]), 0] ################################# classs_index = [1 for i in range(len(gp.BHG_chineseName))] # ¿ª±ÙÒ»¸ölist£¬³¤¶ÈºÍÄ¿±êÊýÁ¿Ò»Ö£¬Ã¿¸öÀïÃæ±£´æµÄÊÇclassµÄË÷Òý for i in range(len(boxes_fine)): # tprint1 = time.time() xyxy = boxes_fine[i][:4] cls = boxes_fine[i][4] index = boxes_fine[i][5] classs_index_show = classs_index[int(cls)] classs_index[int(cls)] += 1 # tprint2 = time.time() # if gp.DebugPrint_flag: # print("tprint1 is ", (tprint2 - tprint1) * 1000) # tcls1 = time.time() b = (((int(xyxy[3]) - int(xyxy[1])) * (int(xyxy[2]) - int(xyxy[0]))) < 15000) #a =(xyxy.item()[3] - xyxy.item()[1]) * (xyxy.item()[2] - xyxy.item()[0]) # tcls2 = time.time() # if gp.DebugPrint_flag: # print("tcls1 is ", (tcls2 - tcls1) * 1000) if cls == 0 or (((int(xyxy[3]) - int(xyxy[1])) * (int(xyxy[2]) - int(xyxy[0]))) < gp.fine_thresh2): #if cls == gp.birdIndex or (torch.sub(torch.mul(torch.sub(xyxy[3], xyxy[1]), torch.sub(xyxy[2], xyxy[0])),gp.fine_thresh2) < 0): label = gp.NoneLabel else: # tcls3 = time.time() label = gp.BHG_chineseName[int(cls)] #label = "{} No.{}".format(label,str(classs_index_show).zfill(3)) spiec_number[int(cls)] += 1 # tcls4 = time.time() # if gp.DebugPrint_flag: # print("tcls2 is ", (tcls4 - tcls3) * 1000) # tprint3 = time.time() # if gp.DebugPrint_flag: # print("tprint2 is ", (tprint3 - tprint2) * 1000) #label = f'{int(cls)} {conf:.2f}' palette = (2 ** 11 - 1, 2 ** 15 - 1, 2 ** 20 - 1) colors = gp.color_box #tuple(int((p * (4 ** 2 - 4 + 1)) % 220) for p in palette) # tprint4 = time.time() # if gp.DebugPrint_flag: # print("tprint3 is ", (tprint4 - tprint3) * 1000) draw = plot_one_box(xyxy, draw, height_or, width_or, label=label, Index=index, color=colors, line_thickness=int(1.5 * gp.radioShowtext)) # tpChin1 = time.time() if num_image > gp.showThresh: gp.num_count_old = len(boxes_fine) gp.spiec_number_old = spiec_number num_image = 0 # tpChin2 = time.time() # if gp.DebugPrint_flag: # print("tpChin1 is ", (tpChin2 - tpChin1) * 1000) #if width_or == 2560: if gp.ShowAllNumFlag: draw = cv2AddChineseText(draw, "识别总数: {}只".format(str(gp.num_count_old)), (50 * gp.radioShowX, 1100 * gp.radioShowY), textColor=gp.color_show, textSize=int(50 * gp.radioShowtext)) # tpChin3 = time.time() # if gp.DebugPrint_flag: # print("tpChin2 is ", (tpChin3 - tpChin2) * 1000) draw = cv2AddChineseText(draw, "分类识别:", (50 * gp.radioShowX, 1150 * gp.radioShowY), textColor=gp.color_show, textSize=int(50 * gp.radioShowtext)) # tpChin4 = time.time() # if gp.DebugPrint_flag: # print("tpChin3 is ", (tpChin4 - tpChin3) * 1000) for i in range(1, len(spiec_number)): if i < 5: draw = cv2AddChineseText(draw, "{}:{}只".format(str(gp.BHG_chineseName[i]), str(gp.spiec_number_old[i])), (300 * gp.radioShowX + (i - 1) * 280 * gp.radioShowX, 1150 * gp.radioShowY), textColor=gp.color_show, textSize=int(50* gp.radioShowtext)) else: draw = cv2AddChineseText(draw, "{}:{}只".format(str(gp.BHG_chineseName[i]), str(gp.spiec_number_old[i])), (50 * gp.radioShowX + (i - 5) * 280 * gp.radioShowX, 1200 * gp.radioShowY), textColor=gp.color_show, textSize=int(50* gp.radioShowtext)) img_show = cv2.cvtColor(np.asarray(img_show), cv2.COLOR_RGB2BGR) gp.img_show = img_show # tpChin2 = time.time() # if gp.DebugPrint_flag: # print("tpChin is ", (tpChin2 - tpChin1) * 1000) #cv2.waitKey(0) # cv2.imwrite("../data_MPKFNet/RTSP/" + "{}.jpg".format(time.strftime('%Y%m%d%H%M%S', time.localtime(time.time()))), # image_or) # cv2.imwrite( # "../data_MPKFNet/RTSP/result/" + "{}.jpg".format(time.strftime('%Y%m%d%H%M%S', time.localtime(time.time()))), # self.img) if gp.DebugPrint_flag: print(self.name + " is finish") num_image += 1 return img_show, num_image def getrestartresult(self, IndexMax): #Ä¿±ê×·×Ù×î´óµÄË÷Òý list = [[] for i in range(IndexMax)] #ÓÃÀ´´æ·Å°´ÕÕIndexË÷ÒýµÄ²»Í¬Í¼Æ¬µÄ£¨bbox(x,y,x,y),label£© for Index in range(1, IndexMax+1): #¶ÔËùÓпÉÄܵÄIndex½øÐбéÀú for imagei in range(5): #¶ÔÎåÕÅͼƬ½á¹û½øÐбéÀú a = gp.Finerestartboxes[imagei] b = gp.Finerestartboxes[imagei][:, 5] c = np.where(b == Index) #ÐèÒªÓÅ»¯ FinerestartboxesindexList = gp.Finerestartboxes[imagei][:, 5].tolist() if Index in FinerestartboxesindexList: RawIndex = FinerestartboxesindexList.index(Index) else: RawIndex = -1 # RawIndex = np.where(gp.Finerestartboxes[imagei][:,5] == Index) #»ñÈ¡µ±Ç°Í¼Æ¬ÖУ¬µ±Ç°IndexËù¶ÔÓ¦µÄÐУ¬Ä¿µÄΪÁË»ñµÃ£¬µ±Ç°Í¼Æ¬Öе±Ç°index±àºÅµÄÄ¿±êµÄ£¨bbox(x,y,x,y),label£© # if (RawIndex[0] == []) | (RawIndex == None): # RawIndex = -1 # else: # RawIndex = RawIndex[0][0] if RawIndex == -1 : #˵Ã÷ûÓÐÕÒµ½ continue else: list[Index-1].append(gp.Finerestartboxes[imagei][RawIndex,-2:]) #½«£¨bbox(x,y,x,y),label,index£© È¡³öÀ´ #µ±Ç°Í¼Æ¬±éÀú½áÊø£¬ÒѾ­ÕÒµ½Index¶ÔÓ¦µÄÄ¿±êÐÅÏ¢ result = [] #ÓÃÀ´´æ·Å×îÖյĽá¹û ½á¹ûΪ¶þά for Index in range(IndexMax):#£¨bbox(x,y,x,y),label,index£©*IndexMax temp = np.array(list[Index]) #»ñÈ¡Index¶ÔÓ¦µÄ¶þάnumpy£¬ÔÚ²»Í¬Í¼Æ¬ÀïÃæµÄÄ¿±êÐÅÏ¢£¨bbox(x,y,x,y),label,index£© print(type(temp)) print(temp) temp1 = temp.mean(axis=0) #¶Ô£¨bbox(x,y,x,y),label,index£©È¡Æ½¾ùÖµ£¬µ«ÊÇlabel²»ÄÜȡƽ¾ùÖµ£¬ÐèÒª¼ÌÐø´¦Àí,È¥µôÁËxyxy temp2 = temp[:, 0] #È¡³öÀ´labelµÄÒ»ÁÐ label_max = statistics.mode(temp2) #Çó³öÀ´³öÏÖlabel×î¶àµÄ½á¹û temp1[0] = label_max #½«label_max Ìæ»»µ½Ô­ÏȵĽá¹ûÖÐ result.append(temp1) return np.array(result) def systemPre(self, img, numPre): height_or, width_or = img.shape[:2] print("image size is ", height_or, width_or) is_ex_bird, labels, boxes, scores = self.DetectionClass.check_object(img) if is_ex_bird: for i in range(len(boxes)): if boxes[i][0] < 0: boxes[i][0] = 0 if boxes[i][1] < 0: boxes[i][1] = 0 if boxes[i][2] < 0: boxes[i][2] = 0 if boxes[i][3] < 0: boxes[i][3] = 0 for i in range(len(labels)): labels[i] = 0 # gp.birdIndex boxes_fine, labels_fine, scores_fine = self.fine_grained.inference(img, boxes, labels, scores, 8) print("System preheating {} is ok".format(str(numPre))) class Density_mul (threading.Thread): def __init__(self, threadID, name, imagetime, DensityClass): threading.Thread.__init__(self) self.threadID = threadID self.name = name self.imagetime = imagetime self.DensityClass = DensityClass def run(self): gp.Density_thread_flag = True if gp.DebugPrint_flag: print(self.name +self.imagetime," is starting") i = 0 while (i<=29): # if(len(gp.ImgDealQueue)<=i): # gp.event.wait() # if ((i+1)%25==0): #˵Ã÷ÊǵÚ25*nÖ¡,¼ä¸ô1s´¦ÀíÒ»´Î if gp.DebugPrint_flag: print("Density is1 ", len(gp.ImgDealQueue), " ", i) print("Density is waiting") if (len(gp.ImgDealQueue) <= i): gp.eventDensity.wait() if gp.DebugPrint_flag: print("Density is2 ", len(gp.ImgDealQueue), " ", i) print("Density is not waiting") self.DensityClass.inference(gp.ImgDealQueue[i],self.imagetime,str(i).zfill(4)) gp.eventDensity.clear() i += 1 gp.Density_thread_flag = False gp.is_ex_bird = False gp.destory_image = True if gp.DebugPrint_flag: print(self.name + self.imagetime, " is over") class Detection_mul (threading.Thread): def __init__(self, threadID, name, imagetime, DetectionClass, Fine_grainedClass): threading.Thread.__init__(self) self.threadID = threadID self.name = name self.imagetime = imagetime self.DetectionClass = DetectionClass self.Fine_grainedClass = Fine_grainedClass def run(self): gp.Detection_thread_flag = True if gp.DebugPrint_flag: print(self.name, self.imagetime, " is starting") i = 0 while (i<=29): if gp.DebugPrint_flag: print("Detection is1 ", len(gp.ImgDealQueue), " ", i) # if (len(gp.ImgDealQueue) <= i): # gp.event.wait() # if ((i+1)%25==0): #˵Ã÷ÊǵÚ25*nÖ¡,¼ä¸ô1s´¦ÀíÒ»´Î print("Detection is waiting") if (len(gp.ImgDealQueue) <= i): gp.eventDetection.wait() if gp.DebugPrint_flag: print("Detection is2 ", len(gp.ImgDealQueue), " ", i) print("Detection is not waiting") self.DetectionClass.inference(gp.ImgDealQueue[i], self.imagetime, str(i).zfill(4), self.Fine_grainedClass) gp.eventDetection.clear() i += 1 gp.Detection_thread_flag = False gp.is_ex_bird = False gp.destory_image = True if gp.DebugPrint_flag: print(self.name, self.imagetime, " is over") class SaveImage_mul (threading.Thread): def __init__(self, threadID, name, imagetime, root): threading.Thread.__init__(self) self.threadID = threadID self.name = name self.imagetime = imagetime self.root = root self.rootpath = os.path.join(self.root, self.imagetime) def run(self): gp.SaveImage_thread_flag = True if not os.path.exists(self.rootpath): os.mkdir(self.rootpath) if not os.path.exists(self.rootpath + "/" + "RawImage" + "/"): os.mkdir(self.rootpath + "/" + "RawImage" + "/") if gp.DebugPrint_flag: print(self.name, self.imagetime, " is starting") i = 0 while (i<=29): # if(len(gp.ImgDealQueue)<=i): # gp.event.wait() # if ((i+1)%25==0): #˵Ã÷ÊǵÚ25*nÖ¡,¼ä¸ô1s´¦ÀíÒ»´Î if gp.DebugPrint_flag: print("Saveimage is1 ", len(gp.ImgDealQueue), " ", i) print("Saveimage is waiting") if (len(gp.ImgDealQueue) <= i): gp.eventSaveimage.wait() if gp.DebugPrint_flag: print("Saveimage is2 ", len(gp.ImgDealQueue), " ", i) print("Saveimage is not waiting") cv2.imwrite(os.path.join(self.rootpath, "RawImage", "{}.jpg".format(str(i).zfill(4))), gp.ImgDealQueue[i]) gp.eventSaveimage.clear() if gp.DebugPrint_flag: print("Saveimage finish once") i += 1 gp.SaveImage_thread_flag = False gp.is_ex_bird = False gp.destory_image = True if gp.DebugPrint_flag: print(self.name, self.imagetime, " is over") class ThreadMonitor_mul (threading.Thread): def __init__(self, threadID, name): threading.Thread.__init__(self) self.threadID = threadID self.name = name def run(self): if gp.DebugPrint_flag: print(self.name, " is starting") while 1: # print(gp.Density_thread_flag,gp.Detection_thread_flag,gp.SaveImage_thread_flag) if (gp.Density_thread_flag) == False and (gp.Detection_thread_flag == False) and (gp.SaveImage_thread_flag == False): #˵Ã÷Èý¸ö½ø³Ì¶¼ÒѾ­½áÊøÁË gp.ImgDealQueue = [] class isopenmul (threading.Thread): def __init__(self, threadID, name, DetectionClass): threading.Thread.__init__(self) self.threadID = threadID self.name = name self.DetectionClass = DetectionClass def run(self): while 1: # print(gp.Density_thread_flag,gp.Detection_thread_flag,gp.SaveImage_thread_flag) if (gp.Vedio_thread_flag == False) and (gp.Density_thread_flag == False) and ( gp.Detection_thread_flag == False) and (gp.SaveImage_thread_flag == False) and ( gp.is_ex_bird == False) and gp.num_detection > 24: gp.num_detection = 0 if gp.DebugPrint_flag: print("check is start") gp.is_ex_bird, labels, boxes, scores = self.DetectionClass.check_object(gp.frame_openmul) if gp.is_ex_bird: if gp.DebugPrint_flag: print(gp.is_ex_bird) gp.image_time = gp.Camera_device + time.strftime('%Y%m%d%H%M%S', time.localtime(time.time())) if gp.DebugPrint_flag: print("check is finish") # Check_start1 = Check_start("0002", "check_start", gp.frame_openmul, self.DetectionClass) # Check_start1.start() if __name__ == '__main__': boxes_list = [[0, 0, 49, 49], [0, 0, 100, 100], [0, 0, 100, 100], [0, 0, 49, 49], [0, 0, 100, 100], [0, 0, 49, 49], [0, 0, 100, 100], [0, 0, 100, 100], [0, 0, 100, 100], [0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100],[0,0,100,100]] label_list = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1] score_list = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1] boxes_list = torch.Tensor(boxes_list) label_list = torch.Tensor(label_list) score_list = torch.Tensor(score_list) path = "/home/aibird/model/data_MPKFNet/testimage/20220324105705.jpg" image = cv2.imread(path) Fine_grained1 = Fine_grained_T(gp.devices, "root") #image = torch.from_numpy(image) boxes_fine, labels_fine, scores_fine = Fine_grained1.inference(image, boxes_list, label_list, score_list, 8) print("ok")