MSCOCO数据集转VOC数据集训练目标检测模型

Images
2014 Train images [83K/13GB]
2014 Val images [41K/6GB]
2014 Test images [41K/6GB]
2015 Test images [81K/12GB]
2017 Train images [118K/18GB]
2017 Val images [5K/1GB]
2017 Test images [41K/6GB]
2017 Unlabeled images [123K/19GB]

Annotations
2014 Train/Val annotations [241MB]
2014 Testing Image info [1MB]
2015 Testing Image info [2MB]
2017 Train/Val annotations [241MB]
2017 Stuff Train/Val annotations [1.1GB]
2017 Panoptic Train/Val annotations [821MB]
2017 Testing Image info [1MB]
2017 Unlabeled Image info [4MB]

我是下载的2017数据集，对应的标注文件如下图所示：

因为是进行目标检测，所以，需要上图中黄色标注的两个文件即可：
打开黄色标记的json文件，数据格式如下所示：

{
    "info": info, # dict
    "licenses": [license], # list ，内部是dict
    "images": [image], # list ，内部是dict
    "annotations": [annotation], # list ，内部是dict
    "categories": # list ，内部是dict
}

字段如下：
info

"info": { # 数据集信息描述
        "description": "COCO 2017 Dataset", # 数据集描述
        "url": "http://cocodataset.org", # 下载地址
        "version": "1.0", # 版本
        "year": 2017, # 年份
        "contributor": "COCO Consortium", # 提供者
        "date_created": "2017/09/01" # 数据创建日期
    },

licenses

"licenses": [
        {
            "url": "http://creativecommons.org/licenses/by-nc-sa/2.0/",
            "id": 1,
            "name": "Attribution-NonCommercial-ShareAlike License"
        },
        ……
        ……
    ],

images

"images": [
        {
            "license": 4,
            "file_name": "000000397133.jpg", # 图片名
            "coco_url":  "http://images.cocodataset.org/val2017/000000397133.jpg",# 网路地址路径
            "height": 427, # 高
            "width": 640, # 宽
            "date_captured": "2013-11-14 17:02:52", # 数据获取日期
            "flickr_url": "http://farm7.staticflickr.com/6116/6255196340_da26cf2c9e_z.jpg",# flickr网路地址
            "id": 397133 # 图片的ID编号（每张图片ID是唯一的）
        },
        ……
        ……
    ],

categories

"categories": [ # 类别描述
        {
            "supercategory": "person", # 主类别
            "id": 1, # 类对应的id （0 默认为背景）
            "name": "person" # 子类别
        },
        {
            "supercategory": "vehicle", 
            "id": 2,
            "name": "bicycle"
        },
        {
            "supercategory": "vehicle",
            "id": 3,
            "name": "car"
        },
        ……
        ……
    ],

注： bicycle 与car都属于vehicle，但两者又属于不同的类别。例如：羊（主类别）分为山羊、绵羊、藏羚羊（子类别）等

annotations

"annotation": [
        {
            "segmentation": [ # 对象的边界点（边界多边形）
                [
                    224.24,297.18,# 第一个点 x,y坐标
                    228.29,297.18, # 第二个点 x,y坐标
                    234.91,298.29,
                    ……
                    ……
                    225.34,297.55
                ]
            ],
            "area": 1481.3806499999994, # 区域面积
            "iscrowd": 0, # 
            "image_id": 397133, # 对应的图片ID（与images中的ID对应）
            "bbox": [217.62,240.54,38.99,57.75], # 定位边框 [x,y,w,h]
            "category_id": 44, # 类别ID（与categories中的ID对应）
            "id": 82445 # 对象ID，因为每一个图像有不止一个对象，所以要对每一个对象编号（每个对象的ID是唯一的）
        },
        ……
        ……
        ]

注意，单个的对象（iscrowd=0)可能需要多个polygon来表示，比如这个对象在图像中被挡住了。而iscrowd=1时（将标注一组对象，比如一群人）的segmentation使用的就是RLE格式。

MSCOCO数据集转VOC数据集代码：

import os, json

jsonPath = r"E:\ai\MSCOCO\annotations\instances_val2017.json"       # json文件存放路径
new_dir=r"E:\ai\MSCOCO\testXml"                                     # 转换得到的xml保存路径


# 转换xml文件
def bboxes2xml(folder, img_name, width, height, gts, xml_save_to):
    xml_file = open((xml_save_to + '/' + img_name + '.xml'), 'w')
    xml_file.write('<annotation>\n')
    xml_file.write('    <folder>' + folder + '</folder>\n')
    xml_file.write('    <filename>' + str(img_name) + '.jpg' + '</filename>\n')
    xml_file.write('    <size>\n')
    xml_file.write('        <width>' + str(width) + '</width>\n')
    xml_file.write('        <height>' + str(height) + '</height>\n')
    xml_file.write('        <depth>3</depth>\n')
    xml_file.write('    </size>\n')

    for gt in gts:
        xml_file.write('    <object>\n')
        xml_file.write('        <name>' + str(gt[0]) + '</name>\n')
        xml_file.write('        <pose>Unspecified</pose>\n')
        xml_file.write('        <truncated>0</truncated>\n')
        xml_file.write('        <difficult>0</difficult>\n')
        xml_file.write('        <bndbox>\n')
        xml_file.write('            <xmin>' + str(gt[1]) + '</xmin>\n')
        xml_file.write('            <ymin>' + str(gt[2]) + '</ymin>\n')
        xml_file.write('            <xmax>' + str(gt[3]) + '</xmax>\n')
        xml_file.write('            <ymax>' + str(gt[4]) + '</ymax>\n')
        xml_file.write('        </bndbox>\n')
        xml_file.write('    </object>\n')

    xml_file.write('</annotation>')
    xml_file.close()


image_bbox = {}  # 图像的id是键名，物体的标注信息为键值

with open(jsonPath, "r") as file:
    read_data = file.read()
loads_dict = json.loads(read_data)
# print(loads_dict.keys())

id_image_dict = {}
for images in loads_dict["images"]:
    temp = []
    temp.append(images["file_name"])
    temp.append(images["height"])
    temp.append(images["width"])

    id_image_dict[images["id"]] = temp

# print(id_image_dict)

id_label_dict = {}  # 物体id类别对照字典
categories_ = loads_dict["categories"]
for categorie in categories_:
    id_label_dict[categorie["id"]] = categorie["name"]
# print(id_label_dict)

annotations = loads_dict["annotations"]  # 注解
for annotation in annotations:
    image_name = annotation["image_id"]  # 对应的图像id
    object_bbox = annotation["bbox"]  # 物体标注狂
    ob_id = annotation["category_id"]  # 物体的类别id
    if image_name not in image_bbox.keys():
        image_bbox[image_name] = []
    object_bbox.insert(0, id_label_dict[ob_id])
    image_bbox[image_name].append(object_bbox)

# print(image_bbox)

for key in image_bbox.keys():
    image_info = id_image_dict[key]  # 得到图像的名称、高度、宽度列表
    image_name = image_info[0]
    image_height = image_info[1]
    image_width = image_info[2]
    gts = []
    for ob_info in image_bbox[key]:
        gt_temp = []
        gt_temp.append(ob_info[0])  # 物体类别
        xmin = round(ob_info[1])
        ymin = round(ob_info[2])

        xmax=xmin+round(ob_info[3])
        ymax=ymin+round(ob_info[4])
        gt_temp=gt_temp+[xmin,ymin,xmax,ymax]
        gts.append(gt_temp)
    folder = "images"
    img_name = image_name.split(".")[0]
    width = image_width
    height = image_height
    xml_save_to = new_dir
    bboxes2xml(folder, img_name, width, height, gts, xml_save_to)
    print("--------------------------------")
    print("done")

labelimag查看数据标注效果：