利用python将数据集按txt文本分类

第二弹 分类test和train
根据stanford数据集的1.6W张图片的txt文本分类test和train，原标签文件为mat格式，可自行搜索转换为txt文本。

import os
import shutil
import cv2

# 读入分类的标签txt文件
#label_file = open("E:\\pythonProject\\data\\train_perfect_preds.txt", 'r')
label_file = open("D:\\Desktop\\three_small\\Stanford-data\\mat2txt.txt", 'r')
# 原始文件的根目录
#input_path = "E:\\pythonProject\\data\\cars_train"
input_path = "D:\\Desktop\\three_small\\Stanford-data\\car_ims"
# 保存文件的根目录
#output_path = "E:\\pythonProject\\result"
output_path = "D:\\Desktop\\three_small\\Stanford-data"
# 标签数组
#lables = ["Classical", "Rock", "Symphony", "Country"]
paths = ['/fdfs_data/data/', '/fdfs_data/data1/']

trainpath = '/home/*/*/SR/raisr/test'

#file_name = "E:\pythonProject\sort\data\cars_train"
file_name = "D:\Desktop\\three_small\Stanford-data\car_ims"  #\t 转义
def get_img_file(file_name):
    imagelist = []
    for parent, dirnames, filenames in os.walk(file_name):
        for filename in filenames:
            if filename.lower().endswith(
                    ('.bmp', '.dib', '.png', '.jpg', '.jpeg', '.pbm', '.pgm', '.ppm', '.tif', '.tiff')):
                imagelist.append(os.path.join(parent, filename))
                #print(filename)
        return imagelist

list = get_img_file(file_name)  #图片的路径数组
#print(list)
#print(file_name)
# 一行行读入标签文件
data = label_file.readlines()
#data1 = pic_file.readlines()
# 计数用
i = 0


for line in data:
    # 通过空格拆分成数组
    str1 = line.split(" ")
    print(str1)
    # 第一个是文件名
    #file_name = str1[0]
    # 第二个是标签类别，并去除最后的换行字符
    #file_label = str1[1].strip()
    file_label = str1[2].strip() #第三个标识是标签
    # 原始文件的路径
    #str2=file_label.zfill(5)
    #old_file_path = os.path.join(input_path, str2)
    old_file_path = list[i]
    # 新文件路径
    new_file_path = ""

    # 如果文件名中有test字符，将其保存至test文件夹下的对应标签文件夹中
    #if "test" in file_name:
    if str1[3] == '0\n':  #分割后的字符包括换换行符
        new_file_path = os.path.join(output_path, "car_ims_train", file_label)
        #print(new_file_path + "\t进if了")
    # 如果文件名中有 train 字符，将其保存至train文件夹下的对应标签文件夹中
    #elif "train" in file_name:
        #new_file_path = os.path.join(output_path, "train", lables[int(file_label) - 1])
    elif str1[3] == '1\n':
        new_file_path = os.path.join(output_path, "car_ims_test", file_label)
    # 如果路径不存在，则创建
    #print(new_file_path)
    if not os.path.exists(new_file_path):
        print("路径 " + new_file_path + " 不存在，正在创建......")
        os.makedirs(new_file_path)

    # 新文件位置
    new_file_path = os.path.join(new_file_path, str(i))
    img = cv2.imread(old_file_path)
    sp = img.shape
    height = sp[0]
    width = sp[1]
    channel = sp[2]
    # 复制文件
    #if height>400 and width >400:
    print("" + str(i) + "\t正在将 " + old_file_path + " 复制到 " + new_file_path)
    shutil.copyfile(old_file_path, new_file_path+'.jpg')
    #else:
    print(old_file_path,sp)
    i = i + 1
# 完成提示
print("完成")