（五）yolov3(c++版)+win训练数据之ImageSets分类

环境 win10 + python3.x + pycharm

本次目的是为了把我们的图片做一下分类，
image_set:图片分类的集合，产生ImageSets/main/ 下面的几个txt文件，训练时依靠这个去匹配相应的图片去训练或测试

1.在Yolov3Train下运行MakeImageSetMainTxt.py，这里VOC后面的数字我的是2025

2.生成对应的txt

3.其实就是图片的序号，训练给后面生成训练label当检索使用

4.附录代码

# -*- coding: utf-8 -*-
# @Author  : xionghao.chen
# @File    : MakeImageSetsMainTxt.py
# @date    : 2020.03.13
# @Software: PyCharm
#产生图片集序列

import os
import random


def make_imagesets(year):
    trainval_percent = 0.2  #训练验证的图片占总训练图片的比例
    train_percent = 0.8     #训练图片占总图片的比例
    xmlfilepath = 'VOCdevkit/VOC%s/Annotations'%(year)
    total_xml = os.listdir(xmlfilepath)

    num = len(total_xml)
    list = range(num)
    tv = int(num * trainval_percent)
    tr = int(tv * train_percent)
    trainval = random.sample(list, tv)
    train = random.sample(trainval, tr)

    ftrainval = open('VOCdevkit/VOC%s/ImageSets/Main/trainval.txt'%(year), 'w')
    ftest = open('VOCdevkit/VOC%s/ImageSets/Main/test.txt'%(year), 'w')
    ftrain = open('VOCdevkit/VOC%s/ImageSets/Main/train.txt'%(year), 'w')
    fval = open('VOCdevkit/VOC%s/ImageSets/Main/val.txt'%(year), 'w')

    for i in list:
        name = total_xml[i][:-4] + '\n'
        if i in trainval:
            ftrainval.write(name)
            if i in train:
                ftest.write(name)
            else:
                fval.write(name)
        else:
            ftrain.write(name)

    ftrainval.close()
    ftrain.close()
    fval.close()
    ftest.close()


if __name__ == '__main__':
    try:
        year = input("input voc year to make_imagesets:")
        make_imagesets(year)
        print('make_imagesets finish')
    except Exception as e:
        print('make_imagesets error', e)

    os.system('pause')