pythonは、トレーニングセットと検証セットを区別するためにピクチャを一括移動します.

#coding=utf-8
import os
import shutil
#https://blog.csdn.net/qikaihuting/article/details/72667740

#folder_capacity = 10000

def copy_files(src_dir, dest_dir):
    if not os.path.exists(dest_dir):
        os.mkdir(dest_dir)

    count = 0
    for item in os.listdir(src_dir):
        abs_item = os.path.join(src_dir, item)
        dst_name = os.path.join(dest_dir, item)
        if os.path.exists(abs_item):
            count +=1
            shutil.copy(abs_item,dst_name)
            
        if count >= 20000:
            break
    print("done")
    return 0
        
def move_file(src_dir,target_dir):
    if not os.path.exists(target_dir):
        os.mkdir(target_dir)
#    count = 0
    for item in os.listdir(src_dir):
        #print(item.split('.')[0])
        int_value = int(item.split('.')[0])
        if int_value%30 == 0:
            #print(int_value)
            src_name = os.path.join(src_dir,item)
   
            target_name = os.path.join(target_dir,item)
#        count +=1
            shutil.move(src_name,target_name)
 
        
#        if count>=20000:
#            break
 
if __name__ =='__main__':
    move_file('E:/semantic_segmentation/models-master/research/deeplab/datasets/yaogan_ADE20K/ADE20K/ADEChallengeData2016/images/training','E:/semantic_segmentation/models-master/research/deeplab/datasets/yaogan_ADE20K/ADE20K/ADEChallengeData2016/images/validation')