Mmdetection trains its own data set -- export coco format of cvat annotation file and related operations

See the previous configuration and the messy problems encountered :https://blog.csdn.net/chenfang0529/article/details/115094036

One 、 export

Use mmdetection Train your own dataset , Dataset use VCAT Annotate , The marked file is a video file , Export image frames and annotation files as COCO Format . There are also PASCAL VOC

Include two files after Export
images and annotations

images Contains image frames

annotations Include dimension files , We only need to modify the third file .

Two 、 Related codes

1. Batch modify the picture name

import os

class BatchRename():

      def rename(self):

          path="D:\\achenf\data\\taxi\\test\\task_2_9_car_test-2021_04_13_13_25_24-coco\images"
          filelist=os.listdir(path)
          total_num = len(filelist)
          i=595
          for item in filelist:
              if item.endswith('.jpg'):
                  src=os.path.join(os.path.abspath(path),item)
                  dst=os.path.join(os.path.abspath(path),''+str(i)+'.jpg') # You can choose the format according to your needs 
                  # dst=os.path.join(os.path.abspath(path),'00000'+format(str(i))+'.jpg') # You can choose the format according to your needs , Custom picture name 
                  try:
                      os.rename(src,dst) #src: old name  dst New name d
                      i+=1
                  except:
                      continue
          print ('total %d to rename & converted %d png'%(total_num,i))

if __name__=='__main__':
    demo = BatchRename()
    demo.rename()

2. Bulk changes json The contents of the document
json in id Etc. need to correspond with the picture .
Needed json It contains five parts ,info,categories,licenses,annotations,images
We just need to modify it annotations and images Two parts .

import json
import os

path = 'D:\\achenf\data\\taxi\\train\\task_2_8_car_test-2021_04_13_13_25_07-coco\\annotations\\test'
dirs = os.listdir(path)

num_flag = 0
for file in dirs: #  Loop to read the file under the path and filter the output 
    if os.path.splitext(file)[1] == ".json": #  Screening csv file 
        num_flag = num_flag +1

        print("path ===== ",file)
        print(os.path.join(path,file))
        with open(os.path.join(path,file),'r') as load_f:
            load_dict = json.load(load_f)
        # print(load_dict)
        # n=len(load_dict["image_id"])
        # print(type(load_dict))
        # for i in load_dict:
        # print(i)

        for i in load_dict['annotations']:
            i['image_id'] = i['image_id'] + 595
            i['id']=i['id']+2032
            # if i['image_id']>=595:
            # i['id']=i['id']+3015
            
    

        for i in load_dict['images']:
            i['id'] = i['id'] + 595
            i['file_name'] = ""+str(i['id'])+".jpg"


        with open(os.path.join(path,file),'w') as dump_f:
            json.dump(load_dict, dump_f)

if(num_flag == 0):
    print(' The selected folder does not exist json file , Please re confirm the folder you want to select ')
else:
    print(' common {} individual json file '.format(num_flag))

Finally, merge the corresponding parts

3、 ... and 、 other

1. analysis xml file , Check the number of annotations in the file

import os
import xml.dom.minidom

res=0
AnnoPath = r'./file_xml/0512/'
Annolist = os.listdir(AnnoPath)
for annotation in Annolist:
    filename =AnnoPath + annotation
    dom = xml.dom.minidom.parse(filename) #  open XML file 
    collection = dom.documentElement #  Get element object 
    objectlist = collection.getElementsByTagName('box') # s
    count = objectlist.length
    res =res+count
    print(" file name ：", filename," Number of dimensions ：", count)
print(" Total labels ：", res)

result ：