垃圾桶满溢识别

针对室外场景大型回收垃圾桶和分类垃圾桶的状态进行识别，如果出现满溢情况，会发生告警。
精度对应场景或者限制条件
1）场景限定：社区、街道和广场等场景
2）图片分辨率：1080p
3）设备/镜头：支持RGB可见光识别；不支持红外识别
4) 识别对象尺寸：最小尺寸为60*80px；遮挡<50%
5）识别的种类：大型回收垃圾桶和分类垃圾桶

This dataset is designed to identify the status of large recycling bins and sorted waste bins in outdoor scenarios, and triggers an alert when an overflow occurs. The accuracy is subject to the following scenarios and constraints:
1) Scenario constraints: scenarios such as communities, streets, and squares
2) Image resolution: 1080p
3) Device/lens requirements: support RGB visible light recognition; infrared recognition is not supported
4) Recognized object size: minimum dimension of 60*80px; occlusion rate <50%
5) Recognized categories: large recycling bins and sorted waste bins

搜集汇总

背景与挑战

背景概述

该数据集是一个算法模型，用于识别室外场景中大型回收垃圾桶和分类垃圾桶的满溢状态，适用于社区、街道和广场等场景，要求图片分辨率为1080p，支持RGB可见光识别，识别对象最小尺寸为60*80px，遮挡需小于50%。