深度学习机器学习常用的脚本-1-数据集转换voc-yolo

voc-yolo数据集转换

该脚本可以很快帮助大家进行数据集格式转换

1. 修改点1见注释@@@@@@@，修改自己数据集的类别
2. 修改文件路径in_file = open(‘VOCdevkit/VOC2007/Annotations/%s.xml’，也可以使用绝对路径

#!/usr/bin/python
# -*- coding:utf-8 -*-

import os
import xml.etree.ElementTree as ET

# classes = ["aeroplane", "bicycle", "bird", "boat", "bottle"]
classes = ['228']  # 第一个要改的地方@@@@@@@@@@自己要测的目标类别


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
    return (x, y, w, h)


def makedirs(path):
    if not os.path.exists(path):
        os.makedirs(path)


def convert_annotation(image_id, out_file):
    in_file = open('VOCdevkit/VOC2007/Annotations/%s.xml' % (image_id), encoding='utf-8')  # （改！）自己的图像标签xml文件的路径
    # out_file = open('VOCdevkit/VOC2007/yolo2_labels/%s.txt' % (image_id), 'w')  # （改！）自己的图像标签txt文件要保存的路径
    tree = ET.parse(in_file)  # 直接解析xml文件
    root = tree.getroot()  # 获取xml文件的根节点
    size = root.find('size')  # 获取指定节点“图像尺寸”
    w = int(size.find('width').text)  # 获取图像宽
    h = int(size.find('height').text)  # 图像高

    for obj in root.iter('object'):
        difficult = obj.find('difficult').text  # xml里的difficult参数
        cls = obj.find('name').text  # 要检测的类别名称name
        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)[0:8] for a in bb]) + '\n')


def generate_txt(txt_name):
    def work(txt_name):
        f = open(os.path.join("VOCdevkit/VOC2007/ImageSets/Main", txt_name), encoding='utf-8')
        for file_name in f.readlines():
            # print(file_name)
            image_id = file_name.split('\n')[0]  # image_id内容类似'0001'
            out_file = open('VOCdevkit/VOC2007/yolo_labels/%s/%s.txt' % (txt_name.split('.')[0], image_id),
                            'w', encoding='utf-8')  # （改！）自己的图像标签txt文件要保存的路径
            convert_annotation(image_id, out_file)
    if txt_name == "test.txt":
        work(txt_name)
    elif txt_name == "train.txt":
        work(txt_name)
    elif txt_name == "trainval.txt":
        work(txt_name)
    elif txt_name == "val.txt":
        work(txt_name)



if __name__ == '__main__':
    # makedirs('VOCdevkit/VOC2007/yolo_labels')
    makedirs('VOCdevkit/VOC2007/yolo_labels/test')
    makedirs('VOCdevkit/VOC2007/yolo_labels/train')
    makedirs('VOCdevkit/VOC2007/yolo_labels/trainval')
    makedirs('VOCdevkit/VOC2007/yolo_labels/val')

    txt_paths = os.listdir("VOCdevkit/VOC2007/ImageSets/Main")
    for txt_name in txt_paths:
        generate_txt(txt_name)