yolov5-7.0分割数据集json、txt文件互转

由于yolo做分割需要txt格式的标签

json->txt：

import os
import cv2
import json
import numpy as np
 
 
def txt_write(x,img_x,img_y,txt):
    data = x['points']
    n = 1
    for x in data:
        for i in x:
            if n % 2 == 0:
                txt.write(' ' + str(round(i/img_x,6)))
                n += 1
            else:
                txt.write(' ' + str(round(i/img_y,6)))
                n += 1  
    txt.write('\n') 
 
 
def json2txt(json_path,save_path):
    txt = open(save_path,'w')
    with open(json_path, "r") as f:
        data = f.read()
    data = json.loads(data)
    img_x = data['imageHeight']
    img_y = data['imageWidth']
    shapes = data['shapes']
 
    for x in shapes:
        #print(x['label'])
        #此处面向不同分类，需要改动下面的标签值，如果是多分类，那么需要增加新的if
        #只是单分类的话，可以直接去掉if，把里面的模块拿出来用
 
        if x['label']=='huahao':
            txt.write('0') 
            txt_write(x,img_x,img_y,txt)
        elif x['label']=='podong':
            txt.write('1') 
            txt_write(x,img_x,img_y,txt)
        elif x['label']=='furong':
            txt.write('2') 
            txt_write(x,img_x,img_y,txt)
        elif x['label']=='fengtou':
            txt.write('3') 
            txt_write(x,img_x,img_y,txt)
        elif x['label']=='zhanwu':
            txt.write('4') 
            txt_write(x,img_x,img_y,txt)
        elif x['label']=='zanting':
            txt.write('5') 
            txt_write(x,img_x,img_y,txt)
        elif x['label']=='guashang':
            txt.write('6') 
            txt_write(x,img_x,img_y,txt)
    txt.close()
    
# # 单文件测试
# save_dir = "" #文件路径
# name = 'test'
# save_path = save_dir + name + '.txt' # 也可以是.doc
# json_path = './zw100-two.json'
# json2txt(json_path,save_path)
 
#文件夹
json_dir = ''
save_dir = ''
files = os.listdir(json_dir)
os.makedirs(save_dir, exist_ok=True)
num = 1
for file in files :
    name = file[0:-5]
    json_path = json_dir + '/' + name + '.json'
    save_path = save_dir + '/' + name + '.txt'
    json2txt(json_path,save_path)
    print(num,'/',len(files),':',name)
    num += 1

如果你的json中有中文

类似这样 会报错 

Traceback (most recent call last): File "d:/code_python/琢磨琢磨/json2txt.py", line 75, in <module> json2txt(json_path,save_path) File "d:/code_python/琢磨琢磨/json2txt.py", line 25, in json2txt data = json.loads(data) File "D:\Ana\envs\tomato\lib\json\__init__.py", line 348, in loads return _default_decoder.decode(s) File "D:\Ana\envs\tomato\lib\json\decoder.py", line 337, in decode obj, end = self.raw_decode(s, idx=_w(s, 0).end()) File "D:\Ana\envs\tomato\lib\json\decoder.py", line 353, in raw_decode obj, end = self.scan_once(s, idx) json.decoder.JSONDecodeError: Invalid \escape: line 39 column 26 (char 718)

这时候我们需要将json中的这段内容删掉

import os
import json

def delete_line_json(json_path):
    # 读取JSON文件
    with open(json_path, 'r', encoding='utf-8') as f:
        data = f.readlines()
    print(data)

    # 删除指定行数据
    print(len(data))
    for i in range(len(data)):
        if "imagePath" in data[i]:
            del data[i]
            break
        

    # 将数据写入文件
    with open(json_path, 'w', encoding='utf-8') as f:
        f.write(''.join(data))

# delete_line_json('./zw100-two.json')#单json文件测试
# 遍历指定目录下所有的JSON文件


directory=""#json文件夹
def batch_delete_line_json(directory):
    for file in os.listdir(directory):
        if file.endswith('.json'):
            file_path = os.path.join(directory, file)
            delete_line_json(file_path)

batch_delete_line_json(directory)

# 调用批量删除函数
# batch_delete_line_json('/path/to/json/folder', 39) # 将39行数据删除

然后再走上面的脚本转txt文件就ok就可以用来做yolo的分割喽

txt转json：

import json
import os
from PIL import Image
import numpy as np
import base64



img_path ="/home/zhh/code_python/txt-json/img/"#图片地址
txt_path ="txt/"#txt地址
json_path="json1/"#json保存地址
labels = os.listdir(txt_path)
#标签名字
class_names = ["vehicle"]


for label in labels:
	#设置头部
    tupian = {
        "version": "5.1.1",
        "flags": {},
        "shapes": [],
        "imagePath": "",
        "imageData": "",
        "imageHeight": "",
        "imageWidth":"" 

    }

    
    #获取图片大小，并保存
    img = np.array(Image.open(img_path+label.strip('.txt') + '.jpg'))
    sh,sw = img.shape[0],img.shape[1] 
    tupian["imageHeight"] = sh
    tupian["imageWidth"] = sw

    tupian["imagePath"]=os.path.basename(img_path)#获取文件名称

    # 获取图片数据并保存为base64编码
    with open(img_path + label.strip('.txt') + '.jpg', "rb") as f:
        image_data = f.read()
    base64_data = base64.b64encode(image_data)
    tupian["imageData"] = str(base64_data, encoding='utf-8')

    #读取base64编码，
    # 读取txt
    with open(txt_path + label, 'r') as f:
        contents = f.readlines()
        shapes = []
        for content in contents:  # 循环每一行内容
            labeldicts = []
            shapess = {'label': "", "points": [], "group_id": None, "shape_type": "polygon", "flags": {}}  # 后缀
            content = content.strip('\n').split()
            shapess['label'] = class_names[int(content[0])]  # 保存label名称
            del content[0]  # 删除第一个元素，保留坐标
            list__ = [content[i:i + 2] for i in range(0, len(content), 2)]  # 修改数组，两个一组[x,y]
            for i in list__:
                x = float(i[0]) * sw
                y = float(i[1]) * sh
                labeldicts.append([x, y])
            shapess["points"] = labeldicts
            shapes.append(shapess)
        tupian["shapes"] = shapes

    # 将字典转换为json字符串
    json_str = json.dumps(tupian, indent=2)

    # 将json字符串写入文件
    json_filename = label.strip('.txt') + '.json'
    json_filepath = os.path.join(json_path, json_filename)
    with open(json_filepath, 'w') as json_file:
        json_file.write(json_str)