城镇基础设备缺陷检测数据集8887张VOC+YOLO格式
数据集格式：Pascal VOC格式+YOLO格式(不包含分割路径的txt文件，仅仅包含jpg图片以及对应的VOC格式xml文件和yolo格式txt文件)
图片数量(jpg文件个数)：8887
标注数量(xml文件个数)：8887
标注数量(txt文件个数)：8887
标注类别数：20
所在github仓库：datasets_sl
标注类别名称(注意yolo格式类别顺序不和这个对应，而以labels文件夹classes.txt为准):["Broken-Cracked-Manholes","Broken-Damaged-Bridges","Broken-Damaged-Buildings","Broken-Damaged-Pipelines","Broken-Damaged-Staircases","Broken-Poles","Broken-glass","Broken-roof","Building-Debris","Cracked-Damaged-Sidewalks","Cracked-Historical-Building","Cracked-Surfaces","Damaged-Railway","Damaged-roads","Faulty-Building-Walls","Faulty-Shipping-Container","Fire-Damaged","Historical-building","Railway-tracks","Skycrapper-Glasses-Cracked"]
每个类别标注的框数：
Broken-Cracked-Manholes 框数 = 1114破损/开裂的井盖
Broken-Damaged-Bridges 框数 = 211破损/受损的桥梁
Broken-Damaged-Buildings 框数 = 5803破损/受损的建筑
Broken-Damaged-Pipelines 框数 = 745破损/受损的管道
Broken-Damaged-Staircases 框数 = 2014破损/受损的楼梯
Broken-Poles 框数 = 1395断裂/破损的电线杆
Broken-glass 框数 = 3630破碎的玻璃
Broken-roof 框数 = 1629破损的屋顶
Building-Debris 框数 = 3796建筑碎片（或建筑残骸）
Cracked-Damaged-Sidewalks 框数 = 421开裂/受损的人行道
Cracked-Historical-Building 框数 = 1833开裂的历史建筑
Cracked-Surfaces 框数 = 4382开裂的表面
Damaged-Railway 框数 = 457受损的铁路
Damaged-roads 框数 = 4124受损的道路
Faulty-Building-Walls 框数 = 1508故障墙体
Faulty-Shipping-Container 框数 = 790故障集装箱
Fire-Damaged 框数 = 575火灾受损（的建筑/设施）
Historical-building 框数 = 2701历史建筑
Railway-tracks 框数 = 1铁路轨道
Skycrapper-Glasses-Cracked 框数 = 149大楼玻璃开裂
总框数：37278
每个类别占有图片数：
Broken-Cracked-Manholes 占有图片数 = 299
Broken-Damaged-Bridges 占有图片数 = 95
Broken-Damaged-Buildings 占有图片数 = 2826
Broken-Damaged-Pipelines 占有图片数 = 210
Broken-Damaged-Staircases 占有图片数 = 297
Broken-Poles 占有图片数 = 730
Broken-glass 占有图片数 = 988
Broken-roof 占有图片数 = 808
Building-Debris 占有图片数 = 2109
Cracked-Damaged-Sidewalks 占有图片数 = 300
Cracked-Historical-Building 占有图片数 = 350
Cracked-Surfaces 占有图片数 = 898
Damaged-Railway 占有图片数 = 148
Damaged-roads 占有图片数 = 398
Faulty-Building-Walls 占有图片数 = 897
Faulty-Shipping-Container 占有图片数 = 199
Fire-Damaged 占有图片数 = 291
Historical-building 占有图片数 = 299
Railway-tracks 占有图片数 = 1
Skycrapper-Glasses-Cracked 占有图片数 = 100
图片分辨率：640x640
使用标注工具：labelImg
数据集是否增强：是
标注规则：对类别进行画矩形框
重要说明：数据集没有划分训练验证测试集需自行划分
特别声明：本数据集不对训练的模型或者权重文件精度作任何保证
标注及图片情况如下：
