Construction Site Material Stacking Encroachment Detection Image Dataset

In construction site safety management, road occupation caused by improper material stacking is a widespread issue affecting traffic flow and safety management. Current manual inspections and management are inefficient, unable to monitor in real-time and address issues promptly. Therefore, there is an urgent need for an efficient automated monitoring solution. This dataset aims to identify and detect the material stacking conditions at construction sites, aiding in intelligent site management. The data is collected through drones and fixed cameras, covering different weather conditions and time periods to ensure diversity and comprehensiveness. Quality control includes multiple rounds of annotation and expert review to ensure accuracy. The annotation team consists of 10 professionals with experience in image processing. Data preprocessing involves techniques such as image denoising, contrast enhancement, and semantic segmentation. Data is stored in JPEG format, accompanied by annotation information in JSON format.