import xml.dom.minidom from tqdm import tqdm from glob import glob import cv2 import os,argparse import numpy as np from jpg import jpg #xml_files = glob('./temp2/*.xml') class_name = ["Anser albifrons" ,"Fulica atra" ,"Leucogeranus leucogeranus" ,"Egretta garzetta","Platalea leucorodia" ,"Mergellus albellus" ,"Grus monacha" ,"Tringa ochropus" ,"Numenius arquata" ,"Anser indicus","Anas zonorhyncha","Ardea cinerea","Mareca strepera" ,"Mareca penelope" ,"Tadorna ferruginea","Ardea alba","Numenius madagascariensis" ,"Botaurus stellaris" ,"Ciconia boyciana" ,"Anser fabalis" ,"Recurvirostra avosetta" ,"Podiceps cristatus","Vanellus vanellus" ,"Aythya fuligula" ,"Tringa erythropus","Calidris alpina" ,"Ciconia nigra" ,"Gallinula chloropus","Porphyrio poliocephalus","Larus ridibundus" ,"Anser cygnoides" ,"Charadrius alexandrinusw" ,"Pluvialis squatarola" ,"Grus grus" ,"Anser anser" ,"Pluvialis fulva" ,"Charadrius dubius" ,"Anas crecca","Anas platyrhynchos","Spatula clypeata","Phalacrocorax carbo","Mergus merganser" ,"Tadorna tadorna" ,"Tringa nebulariaw","Bucephala clangula" ,"Gallinago gallinago" ,"Anser erythropus","Cygnus columbianus" ,"Ichthyaetus ichthyaetus","Anas acuta" ,"Ardea intermedia"] def cv2_crop(im, box): '''cv2实现类似PIL的裁剪 :param im: 加载好的图像 :param box: 裁剪的矩形,元组(left, upper, right, lower). ''' return im.copy()[box[1]:box[3], box[0]:box[2], :] def convert(size, box): dw = 1./size[0] dh = 1./size[1] x = (box[0] + box[1])/2.0 y = (box[2] + box[3])/2.0 w = box[1] - box[0] h = box[3] - box[2] x = x*dw w = w*dw y = y*dh h = h*dh return (x,y,w,h) def VocToFineWithLabel(num_count,root_path,image_T): xml_files = glob(os.path.join(root_path,"*.xml")) jpg(os.path.join(root_path.replace("Annotations","JPEGImages"))) txt_lists = [] for xml_file in tqdm(xml_files): #print(xml_file+" is doing") DOMTree = xml.dom.minidom.parse(xml_file) collection = DOMTree.documentElement #img_name = collection.getElementsByTagName("filename")[0].childNodes[0].data img_name = xml_file.replace("Annotations","JPEGImages").replace(".xml",".jpg") objects = collection.getElementsByTagName("object") #img = cv2.imread(img_name) img = cv2.imdecode(np.fromfile(img_name,dtype=np.uint8),-1) #cv2.imwrite(img_name,img) #img = cv2.imdecode(np.fromfile(img_name,dtype=np.uint8),-1) # im = Image.open(img_name) # im.save(img_name, quality=100) # im = Image.open(img_name) for object_id in objects: xmin = object_id.getElementsByTagName('bndbox')[0].getElementsByTagName('xmin')[0].childNodes[0].data ymin = object_id.getElementsByTagName('bndbox')[0].getElementsByTagName('ymin')[0].childNodes[0].data xmax = object_id.getElementsByTagName('bndbox')[0].getElementsByTagName('xmax')[0].childNodes[0].data ymax = object_id.getElementsByTagName('bndbox')[0].getElementsByTagName('ymax')[0].childNodes[0].data name = "bird"#object_id.getElementsByTagName('name')[0].childNodes[0].data ratio_x = (3840.0/3840.0) ratio_y = (2160.0/2160.0) #print(ratio_x," ",ratio_y) #print(type(xmin)) #exit() xmin = int(float(xmin)*ratio_x) xmax = int(float(xmax)*ratio_x) ymin = int(float(ymin)*ratio_y) ymax = int(float(ymax)*ratio_y) if ((xmax - xmin)