[数据集][目标检测]垃圾目标检测数据集VOC格式14963张44类别

数据集格式:Pascal VOC格式(不包含分割的txt文件,仅仅包含jpg图片和对应的xml)

图片数量(jpg文件个数):14963

标注数量(xml文件个数):14963

标注类别数:44

标注类别名称:["toiletries","plastic utensils","seasoning bottles","leftovers","chopsticks","ceramic utensils","pots","metal utensils","cutting boards","old clothes","expired drugs","fish bones","metal food cans","peels and pulp","cartons and cartons","plug wires","plastic clothes hangers","vegetable leaves","glassware","tea dregs","shoes","disposable fast food boxes","eggshells","plush toys","metal kitchenware","power bank","dry batteries","pull cans","plastic toys","garbage cans","Wine bottle","defaced paper","ointment","big bone","express paper bag","defaced plastic","edible oil barrel","cigarette butts","beverage bottle","bag","flower pot","pillow","book paper","toothpick"]

对应中文名称:

"洗护用品", "塑料器皿", "调料瓶", "剩饭剩菜", "筷子", "陶瓷器皿", "锅", "金属器皿", "砧板", "旧衣服", "过期药物", "鱼骨", "金属食品罐", "果皮果肉", "纸盒纸箱", "插头电线", "塑料衣架", "菜帮菜叶", "玻璃器皿", "茶叶渣", "鞋", "一次性快餐盒", "蛋壳", "毛绒玩具", "金属厨具", "充电宝", "干电池", "易拉罐", "塑料玩具", "垃圾桶","酒瓶", "污损用纸", "软膏", "大骨头", "快递纸袋", "污损塑料", "食用油桶", "烟蒂", "饮料瓶", "包", "花盆", "枕头", "书籍纸张", "牙签"

每个类别标注的框数:

toiletries count = 1030

plastic utensils count = 666

seasoning bottles count = 543

leftovers count = 1414

chopsticks count = 817

ceramic utensils count = 2662

pots count = 530

metal utensils count = 341

cutting boards count = 466

old clothes count = 394

expired drugs count = 710

fish bones count = 525

metal food cans count = 400

peels and pulp count = 1052

cartons and cartons count = 623

plug wires count = 1277

plastic clothes hangers count = 325

vegetable leaves count = 984

glassware count = 878

tea dregs count = 409

shoes count = 463

disposable fast food boxes count = 329

eggshells count = 458

plush toys count = 655

metal kitchenware count = 284

power bank count = 410

dry batteries count = 445

pull cans count = 414

plastic toys count = 439

garbage cans count = 108

Wine bottle count = 346

defaced paper count = 227

ointment count = 418

big bone count = 413

express paper bag count = 248

defaced plastic count = 1237

edible oil barrel count = 446

cigarette butts count = 433

beverage bottle count = 371

bag count = 442

flower pot count = 211

pillow count = 382

book paper count = 206

toothpick count = 156

使用标注工具:labelImg

标注规则:对类别进行画矩形框

重要说明:类别全是英文不用担心标签中文问题,且数据集经过标注校验保证无错误标注

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,数据集只提供准确且合理标注

下载地址:

https://download.csdn.net/download/FL1623863129/85682652

相关推荐
GISer_Jing5 小时前
AI自动化工作流:智能驱动未来(升级研究生项目!!!)
人工智能·前端框架·自动化
草捏子5 小时前
Agent Skills:让 AI 一次学会、永远记住的能力扩展方案
人工智能
NocoBase6 小时前
【2.0 教程】第 1 章:认识 NocoBase ,5 分钟跑起来
数据库·人工智能·开源·github·无代码
后端小肥肠6 小时前
OpenClaw实战|从识图到公众号内容自动化,我跑通了完整链路
人工智能·aigc·agent
Elastic 中国社区官方博客6 小时前
快速 vs. 准确:衡量量化向量搜索的召回率
大数据·人工智能·elasticsearch·搜索引擎·ai·全文检索
qq_381338506 小时前
【技术日报】2026-03-18 AI 领域重磅速递
大数据·人工智能
NocoBase6 小时前
开源项目管理工具选型指南(2026年最新)
人工智能·开源·无代码
feasibility.6 小时前
AI 爬虫高手养成:Openclaw+Scrapling 手动部署 + 采集策略(以Walmart 电商平台为例)
人工智能·爬虫·科技·机器人·agi·openclaw·scrapling
程序员老猫6 小时前
前端菜鸡狂喜!DeepSeek+Gemini,嘴炮出完整博客方案
人工智能
AI周红伟7 小时前
周红伟:《OpenClaw安全防控:OpenClaw++Skills智能体安全部署、实操和企业应用实操》
人工智能·阿里云·云计算·腾讯云·openclaw