coco json转xml格式(开箱即用代码)

提示：文章写完后，目录可以自动生成，如何生成可参考右边的帮助文档

---

前言

最近使用很多其它模型，大多模型都有自己对应格式，有时其它模型的cocojson格式的数据测试也都要进行数据格式转换，有时候也想看coco json注释显示在图上。那么，你可能需要一个coco json格式转xml格式代码，帮助查看，且我总是会把这些工具代码莫名丢失，重头写较为麻烦。为克服这些问题，本文将使用coco数据集格式作为标准，记录如何将coco数据格式转为xml格式。

---

一、整体逻辑

1、main函数

下载好数据给定json格式路径变量值json_path，指定输出文件路径save_path,然后直接调用cocojson2xml函数，代码如下：

if __name__ == '__main__':
    json_path=r"C:\Users\Administrator\Desktop\Data\coco2017\annotations\train.json"
    save_path=r'C:\Users\Administrator\Desktop\Data\coco2017\123'
    cocojson2xml(json_path, save_path)

2、逻辑

代码自动读取coco json格式images、annotation、categories，并通过image_id和category_id找到每个图对应所有box与对应box的类别，同时从coco json文件解析图像高宽，而后使用函数等逻辑转为xml格式，代码会自动完成。

二、函数介绍

build_dir：构建文件夹
read_json：读取coco json文件注释
get_name：处理字符串路径，去除万一是个路径格式
product_xml：生成路径
cocojson2xml：这是主要集成函数，coco转xml

三、整体代码

介于代码主要为逻辑，而非代码难度，我直接粘贴整个代码，可直接复制使用，代码如下：

import os
import json
from tqdm import tqdm

from lxml.etree import Element, SubElement, tostring, ElementTree
from xml.dom.minidom import parseString

import numpy as np

def build_dir(root):
    import os
    if not os.path.exists(root):
        os.makedirs(root)
    return root
def read_json(json_root):
    with open(json_root, encoding='utf-8') as f:
        json_info = json.load(f)
    return json_info

def get_name(name_str):
    name_str=name_str.split('/')[-1]  # 处理服务器方式
    name_str=name_str.split('\\')[-1]  # 处理windows方式
    return name_str

def product_xml(name_img, boxes, codes, img=None, wh=None,save_path=None):
    '''
    :param img: 以读好的图片
    :param name_img: 图片名字
    :param boxes: box为列表
    :param codes: 为列表
    :return:
    '''
    if img is not None:
        width = img.shape[0]
        height = img.shape[1]
    else:
        assert wh is not None
        width = wh[0]
        height = wh[1]
    # print('xml w:{} h:{}'.format(width,height))

    node_root = Element('annotation')
    node_folder = SubElement(node_root, 'folder')
    node_folder.text = 'VOC2007'

    node_filename = SubElement(node_root, 'filename')
    node_filename.text = name_img  # 图片名字

    node_size = SubElement(node_root, 'size')
    node_width = SubElement(node_size, 'width')
    node_width.text = str(width)

    node_height = SubElement(node_size, 'height')
    node_height.text = str(height)

    node_depth = SubElement(node_size, 'depth')
    node_depth.text = '3'

    for i, code in enumerate(codes):
        box = [boxes[i][0], boxes[i][1], boxes[i][2], boxes[i][3]]
        node_object = SubElement(node_root, 'object')
        node_name = SubElement(node_object, 'name')
        node_name.text = code
        node_difficult = SubElement(node_object, 'difficult')
        node_difficult.text = '0'
        node_bndbox = SubElement(node_object, 'bndbox')
        node_xmin = SubElement(node_bndbox, 'xmin')
        node_xmin.text = str(int(box[0]))
        node_ymin = SubElement(node_bndbox, 'ymin')
        node_ymin.text = str(int(box[1]))
        node_xmax = SubElement(node_bndbox, 'xmax')
        node_xmax.text = str(int(box[2]))
        node_ymax = SubElement(node_bndbox, 'ymax')
        node_ymax.text = str(int(box[3]))

    xml = tostring(node_root, pretty_print=True)  # 格式化显示，该换行的换行
    dom = parseString(xml)

    name = name_img[:-4] + '.xml'

    tree = ElementTree(node_root)

    from lxml import etree

    # Assuming you have the root node already defined as node_root
    # Format the XML and save to a file
    if save_path is not None:
        with open(os.path.join(save_path,name), "wb") as f:
            f.write(tostring(node_root, pretty_print=True))

    # print('name:{},dom:{}'.format(name, dom))
    return tree, name



def cocojson2xml(json_path,save_path):

    json_info = read_json(json_path)
    save_path = build_dir(save_path)
    info_img = json_info['images']
    info_ann = json_info['annotations']
    info_cat = json_info['categories']

    img_id = np.array([info['id'] for info in info_img])
    ann_imgid = np.array([info['image_id'] for info in info_ann])

    ann_count = len(ann_imgid)
    cat_id = [info['id'] for info in info_cat]

    print('loading json file , json convert to dict format.............\n')
    xml_info = {}
    for i, id in tqdm(enumerate(img_id)):
        img_ = info_img[i]
        file_name = img_['file_name']
        w_h = [img_['width'], img_['height']]
        ann_ = [info_ann[j] for j in np.arange(ann_count)[ann_imgid==id]]
        box_lst = [[a['bbox'][0],a['bbox'][1],a['bbox'][0]+a['bbox'][2],a['bbox'][1]+a['bbox'][3] ]
                   for a in ann_]  # 将cocojson左上角转与wh转为xml的左上角与右下角坐标



        catid_lst = [a['category_id'] for a in ann_]
        cat_lst = [info_cat[cat_id.index(c)]['name'] for c in catid_lst]

        xml_info[id] = {'file_name':file_name,'w_h':w_h,'box_lst':box_lst, 'cat_lst':cat_lst}

    print('\ngenerate xml file , save xml file.............')
    for k, v in tqdm(xml_info.items()):
        file_name = v['file_name']
        w_h = v['w_h']
        box_lst = v['box_lst']
        cat_lst = v['cat_lst']
        name_img = get_name(file_name)
        tree, xml_name = product_xml(name_img, box_lst, cat_lst,  wh=w_h,save_path=save_path)

        # tree.write(os.path.join(save_path, xml_name))



if __name__ == '__main__':
    json_path=r"C:\Users\Administrator\Desktop\Data\coco2017\annotations\train.json"
    save_path=r'C:\Users\Administrator\Desktop\Data\coco2017\123'
    cocojson2xml(json_path, save_path)

---

总结

coco json格式转xml格式，复制即可使用处理工具。