yolov5数据集制作 数据集分割+voc格式转txt格式代码

先上数据集分割代码 网上没找到合适的 自己写了一个 分割比例为0.6 ：0.2 ：0.2

import os
import random

oriImgDir = "./change14img"
oriXmlDir = "./change14xml"    #源文件夹
desTestImgDir = "./insect/images/test"   #目标文件夹
desValImgDir = "./insect/images/val"
desTrainImgDir = "./insect/images/train"
desTestXmlDir = "./insect/labels/test"
desValXmlDir = "./insect/labels/val"
desTrainXmlDir = "./insect/labels/train"
# 获取目录下文件名清单
flist = os.listdir(oriImgDir)
random.shuffle(flist)
length = len(flist)
val = int(length * 0.2)
test = int(length * 0.2)

count = 0

while (count < length):
    # 移动测试集
    while test:
        item = flist[count]
        fname = os.path.splitext(item)[0]  # 获得不带后缀的文件名
        os.rename(oriImgDir + "/" + item, desTestImgDir+"/14"+str(count)+".jpg")  # 图片
        os.rename(oriXmlDir + "/" + fname + ".xml", desTestXmlDir+"/14"+str(count)+".xml")
        test -= 1
        count += 1
    # 移动验证集
    while val:
        item = flist[count]
        fname = os.path.splitext(item)[0]  # 获得不带后缀的文件名
        os.rename(oriImgDir + "/" + item, desValImgDir+"/14"+str(count)+".jpg")  # 图片
        os.rename(oriXmlDir + "/" + fname + ".xml", desValXmlDir+"/14"+str(count)+".xml")
        val -= 1
        count += 1
    # 剩下的移到训练集
    item = flist[count]
    fname = os.path.splitext(item)[0]
    os.rename(oriImgDir + "/" + item, desTrainImgDir+"/14"+str(count)+".jpg")  # 图片
    os.rename(oriXmlDir + "/" + fname + ".xml", desTrainXmlDir+"/14"+str(count)+".xml")
    count += 1

比例可以自己改，不需要验证集的话把第二个while循环去掉即可
os.rename的第二个参数是移动过去之后的文件名，可以改，如果源文件夹有很多个且有重名的情况，第二个参数就需要设计成让移过去之后新的文件名不会重复
os函数不支持跨磁盘移动文件

voc标签转txt格式代码 网上找的 源代码地址：
https://github.com/DLLXW/objectDetectionDatasets

'''
2020/6/15,标注文件转换xml转txt（vol to yolo）转完后需添加labels文件，即数字序号对应的标签名。

'''

import xml.etree.ElementTree as ET
import pickle
import os
from os import listdir, getcwd
from os.path import join

classes = ['0', '1', '2', '3', '4', '5', '6', '7', '9', '10', '12', '13', '14']


def convert(size, box):
    dw = 1./(size[0])
    dh = 1./(size[1])
    x = (box[0] + box[1])/2.0 - 1
    y = (box[2] + box[3])/2.0 - 1
    w = box[1] - box[0]
    h = box[3] - box[2]
    x = x*dw
    w = w*dw
    y = y*dh
    h = h*dh
    if w>=1:
        w=0.99
    if h>=1:
        h=0.99
    return (x,y,w,h)

def convert_annotation(rootpath,xmlname):
    xmlfile = os.path.join(xmlpath,xmlname)
    with open(xmlfile, "r", encoding='UTF-8') as in_file:
        txtname = xmlname[:-4]+'.txt'
        print(txtname)
        txtpath = 'yolov5txt/train'#生成的.txt文件会被保存在这里
        if not os.path.exists(txtpath):
            os.makedirs(txtpath)
        txtfile = os.path.join(txtpath,txtname)
        with open(txtfile, "w+" ,encoding='UTF-8') as out_file:
            tree=ET.parse(in_file)
            root = tree.getroot()
            size = root.find('size')
            w = int(size.find('width').text)
            h = int(size.find('height').text)
            out_file.truncate()
            for obj in root.iter('object'):
                difficult = obj.find('difficult').text
                cls = obj.find('name').text
                if cls not in classes or int(difficult)==1:
                    continue
                cls_id = classes.index(cls)
                xmlbox = obj.find('bndbox')
                b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text), float(xmlbox.find('ymin').text), float(xmlbox.find('ymax').text))
                bb = convert((w,h), b)
                out_file.write(str(cls_id) + " " + " ".join([str(a) for a in bb]) + '\n')


if __name__ == "__main__":
    xmlpath='insect/labels/train'
    list=os.listdir(xmlpath)
    for i in range(0,len(list)) :
        path = os.path.join(xmlpath,list[i])
        if ('.xml' in path)or('.XML' in path):
            convert_annotation(xmlpath,list[i])
            print('done', i)
        else:
            print('not xml file',i)

class列表里放的是类别，改成你自己的类别名称，类别如果没添加进去的话就会出现转化出来的txt是空的的情况