数据堂—12,352张搬运物体采集标注数据

12,352张搬运物体采集标注数据，该数据集用于物体搬运检测，包含多个场景，采集角度为俯视，对画面中出现的行人都进行了标注，该数据集可用于对园区、仓库、街道等监控场景中识别出搬运物体的人

A total of 12,352 collected and annotated images for object handling tasks. This dataset is designed for object handling detection, covers multiple scenarios and is captured from a top-down view. All pedestrians appearing in the frames have been annotated. It can be used to identify individuals carrying objects in surveillance scenarios such as campuses, warehouses, streets and other similar scenarios.

搜集汇总

背景与挑战

背景概述

该数据集包含12,352张图像，用于在公园、仓库、街道等监控场景中检测携带物体的行人，采用俯视视角，所有行人都有标注。数据收集于街道、住宅区等多场景，使用监控摄像头在白天和夜晚拍摄，格式包括.jpg、.json和.xml。