first
This commit is contained in:
Semi212 committed
1 parent
54ccfb11fe
commit
a0818377bb
194 files changed
+41681
-262
No files matched your search
+21
-25
@@ -6,19 +6,11 @@ import os
|
||||
from PIL import Image
|
||||
import numpy as np
|
||||
from jpg import jpg
|
||||
txt_root_path = "E:\\dataset\\VOC5"
|
||||
txt_root_path = "/data/DataManagement/Project/005_kunmingdianchi/0002_20220412_800_200/Dataset/VOC0050201"
|
||||
#xml_files = glob('./temp2/*.xml')
|
||||
f_num = open(txt_root_path+"/num.txt", "r")
|
||||
|
||||
num_count = f_num.read()
|
||||
f_num.close()
|
||||
if num_count == "":
|
||||
num_count=0
|
||||
else:
|
||||
num_count= int(num_count)
|
||||
num = 0
|
||||
root_path = "E:\\dataset\\VOC5\\Annotations"
|
||||
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"]
|
||||
image_T = 50
|
||||
num_count = 0
|
||||
root_path = "/data/DataManagement/Project/005_kunmingdianchi/0002_20220412_800_200/Dataset/VOC0050201/Annotations/"
|
||||
def cv2_crop(im, box):
|
||||
'''cv2实现类似PIL的裁剪
|
||||
|
||||
@@ -91,17 +83,17 @@ def convert(size, box):
|
||||
# print(num_count)
|
||||
|
||||
xml_files = glob(os.path.join(root_path,"*.xml"))
|
||||
print(os.path.join(root_path.replace("Annotations","JPEGOriginal")))
|
||||
jpg(os.path.join(root_path.replace("Annotations","JPEGOriginal")))
|
||||
#print(os.path.join(root_path.replace("Annotations","JPEGImages")))
|
||||
jpg(os.path.join(root_path.replace("Annotations","JPEGImages")))
|
||||
f = open(txt_root_path+"/Hatcher_ALL.txt", "a")
|
||||
txt_lists = []
|
||||
for xml_file in tqdm(xml_files):
|
||||
|
||||
print(xml_file+" is doing")
|
||||
#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\\","\\JPEGOriginal\\").replace(".xml",".jpg")
|
||||
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)
|
||||
@@ -116,19 +108,23 @@ for xml_file in tqdm(xml_files):
|
||||
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 = object_id.getElementsByTagName('name')[0].childNodes[0].data
|
||||
ratio_x = (3840.0/1024.0)
|
||||
ratio_y = (2160.0/768.0)
|
||||
print(ratio_x," ",ratio_y)
|
||||
print(type(xmin))
|
||||
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)
|
||||
class_index = class_name.index(str(name).replace("\t","").strip())
|
||||
if ((xmax - xmin)<image_T) | ((ymax - ymin)<image_T):
|
||||
continue
|
||||
#class_index = class_name.index(str(name).replace("\t","").strip())
|
||||
img1 = cv2_crop(img,(xmin, ymin,xmax, ymax))
|
||||
cv2.imwrite("E:\\dataset\\VOC5\Hatcher_One\\" + str(name).replace("\t","").strip().replace(" ","_" )+"_{}.jpg".format(num_count), img1)
|
||||
txt_lists.append(str(name).replace("\t","").strip().replace(" ","_" )+"_{}.jpg".format(num_count)+" "+str(class_index)+'\n')
|
||||
if not os.path.exists(os.path.join(root_path.replace("Annotations","FineDataset_{}".format(str(image_T))),str(name).replace("\t","").strip())):
|
||||
os.makedirs(os.path.join(root_path.replace("Annotations","FineDataset_{}".format(str(image_T))),str(name).replace("\t","").strip()))
|
||||
cv2.imwrite(os.path.join(root_path.replace("Annotations","FineDataset_{}".format(str(image_T))),str(name).replace("\t","").strip(),"{}.jpg".format(str(num_count).zfill(5))), img1)
|
||||
#txt_lists.append(str(name).replace("\t","").strip()+"{}.jpg".format(str(num_count).zfill(5))+" "+str(class_index)+'\n')
|
||||
#f.write(str(name).replace("\t","").strip().replace(" ","_" )+"_{}.jpg".format(num_count)+" "+str(class_index)+'\n')
|
||||
num_count += 1
|
||||
for txt_list in txt_lists:
|
||||
@@ -138,8 +134,8 @@ f_num = open("./num.txt", "w")
|
||||
f_num.write(str(num_count))
|
||||
f_num.close()
|
||||
print(num_count)
|
||||
for xml_file in tqdm(xml_files):
|
||||
os.remove(xml_file)
|
||||
# for xml_file in tqdm(xml_files):
|
||||
# os.remove(xml_file)
|
||||
|
||||
#os.remove(xml_file)
|
||||
print(num_count)
|
||||
|
||||
Reference in new issue
Block a user