json文件批量转换xml

在参加比赛时，官方给出的标记文件是json格式，自己编写代码使其转换为xml格式。使用的是Python 标准库之 xml.etree.ElementTree。

官方给出的格式：

转换代码如下：

import os, json
import copy

import cv2
import numpy as np
try:
    import xml.etree.cElementTree as ElementTree
except ImportError:
    import xml.etree.ElementTree as ElementTree


# <Element 'data' at 0x0000021AE8F79540>  type:<class 'xml.etree.ElementTree.Element'>

template_file = '../intelligent/1.xml'  # xml模板
target_dir = r'./xmll/'   # 保存路径
image_dir = r'./picture/'  # 图片文件夹

train_file = '../intelligent/labels/'   # 存储了图片信息的json文件
file_nameee = ''
json_path = './labels'
# 提取json
def parse_json(d):
    arr = np.array([
        d['image_name'], d['ignore'],d['category_name'],d['id'],d['occlusion'],d['truncation'],d['bbox']
    ])

    return arr



def chinese_english(str):
    if str == 'suv':
        str = 'suv'

    elif str == '专业作业车':
        str = 'Professional_work_car'
    elif str == '儿童':
        str = 'children'
    elif str == '面包车':
        str = 'Van_minibus'
    elif str == '大客车':
        str = 'bus'
    elif str == '大货车':
        str = 'Big_truck'
    elif str == '小货车':
        str = 'buggy'

    elif str == '成年人':
        str = 'adult'
    elif str == '电动/摩托三轮车':
        str = 'Electric_tricycle'
    elif str == '电动/摩托车':
        str = 'Electric_motorcycle'

    elif str == '自行车':
        str = 'bcycle'
    elif str == '轿车':
        str = 'car'
    else:
        str = 'other'

    return str


def read_xml(in_path):
    '''''读取并解析xml文件
       in_path: xml路径
       return: ElementTree'''
    print(in_path)


    #prese()解析xml文件
    tree = ElementTree.parse(in_path)
    return tree
def tiqu(strrr):
    new_str = ""  # 创建一个空字符串
    for ch in strrr:
        if ch.isdigit():  # 字符串中的方法，可以直接判断ch是否是数字
            new_str += ch
        else:
            new_str += " "
    sub_list = new_str.split()  # 对新的字符串切片
    num_list = list(map(int, sub_list))  # map方法，使列表中的元素按照指定方式转变
    return num_list

dir = os.listdir(json_path)
for file in dir:
    trainfiles = json.load(open(train_file+file,encoding='UTF-8'))
    len_nn = len(trainfiles)
    #tree = ElementTree()
    for k, line in enumerate(trainfiles):
        arr = parse_json(line)
        file_name = arr[0]  # 文件名
        print(file_name)




        file_nameee = file_name

        if k == 0:
            print(11)
            label = arr[2]  # 标签名
            # 坐标
            label = chinese_english(label)
            xmin = str(int(arr[6][0] - arr[6][2] / 2))
            xmax = str(int(arr[6][0] + arr[6][2] / 2))  #
            ymin = str(int(arr[6][1] - arr[6][3] / 2))
            ymax = str(int(arr[6][1] + arr[6][3] / 2))
            tree = read_xml(template_file)  # 解析树
            root = tree.getroot()  # 根节点
            root.find('filename').text = file_name
            # size
            sz = root.find('size')
            imggg = image_dir + file_name
            im = cv2.imread(imggg)  # 读取图片信息
            str1 = arr[4]
            str2 = arr[5]
            tiqu_str1 = tiqu(str1)
            tiqu_str2 = tiqu(str2)

            print(tiqu_str1)
            print(tiqu_str2)


            sz.find('height').text = str(im.shape[0])
            sz.find('width').text = str(im.shape[1])
            sz.find('depth').text = str(im.shape[2])
            # object
            obj = root.find('object')
            obj.find('name').text = label
            if tiqu_str1[0] == 0 and tiqu_str2[0] == 0:
                obj.find('truncated').text = str(0)
            else:
                obj.find('truncated').text = str(1)
            bb = obj.find('bndbox')
            bb.find('xmin').text = xmin
            bb.find('ymin').text = ymin
            bb.find('xmax').text = xmax
            bb.find('ymax').text = ymax

        # 有多个标签需要添加object
        else:
            label = arr[2]
            label = chinese_english(label)
            xmin = str(int(arr[6][0] - arr[6][2] / 2))
            xmax = str(int(arr[6][0] + arr[6][2] / 2))  #
            ymin = str(int(arr[6][1] - arr[6][3] / 2))
            ymax = str(int(arr[6][1] + arr[6][3] / 2))

            obj_ori = root.find('object')

            obj = copy.deepcopy(obj_ori)  # 注意这里深拷贝
            str1 = arr[4]
            str2 = arr[5]
            tiqu_str1 = tiqu(str1)
            tiqu_str2 = tiqu(str2)


            if tiqu_str1[0] == 0 and tiqu_str2[0] == 0:
                obj.find('truncated').text = str(0)
            else:
                obj.find('truncated').text = str(1)

            obj.find('name').text = label
            bb = obj.find('bndbox')
            bb.find('xmin').text = xmin
            bb.find('ymin').text = ymin
            bb.find('xmax').text = xmax
            bb.find('ymax').text = ymax
            root.append(obj)

    xml_file = file_nameee.replace('jpg','xml')
    print("deakpf"+xml_file)
    tree.write(target_dir + xml_file, encoding='utf-8')