筛选数据的工具_jishuqianjin的博客_数据筛选工具

一、利用 shutil.copy 进行处理

# -*- coding: utf-8 -*-
import os
import glob
import cv2
import json
import shutil

def move_src_accord_filter(filter_path, src_path, dataset_path):
    files = glob.glob(filter_path + '/*.png')
    count = 0
    test_path = "/home/user/lcw/RFBnet_Data/obqFind_wanli/test_dataset"
    for image_path in files:
        image_name = image_path[image_path.rfind("/")+1: ]
        src_image_path = os.path.join(src_path, image_name)
        if not os.path.exists(src_image_path):
            continue

        json_name = image_name[:image_name.rfind(".")] + ".json"
        json_image_path = os.path.join(src_path, json_name)
        print("json_image_path is: ", json_image_path)
        if count < 16:
            shutil.copy(src_image_path, dataset_path)
            shutil.copy(json_image_path, dataset_path)
        else:
            return
            # shutil.copy(src_image_path, test_path)
            # shutil.copy(json_image_path, test_path)
        count = count + 1
        #shutil.move(json_image_path, dataset_path)



if __name__ == "__main__":
    # filter_path = "/media/Data/wangjunjie/active_learning/test_lcw"
    # src_path = "/home/user/lcw/RFBnet_Data/obqFind_wanli/verify_data"
    # dataset_path ="/media/Data/wangjunjie/active_learning/test_lcw_src"
    filter_path = "/media/newData/paperData/opqData/data_28_filter/verify_result"
    src_path = "/media/newData/paperData/opqData/data_28_filter/verify_src"
    dataset_path = "/home/user/lcw/RFBnet_Data/obqFind_wanli/dataset_lcw"
    if not os.path.exists(dataset_path):
        os.mkdir(dataset_path)
    move_src_accord_filter(filter_path, src_path, dataset_path)# -*- coding: utf-8