Subway Platform Boundary Violation Detection Image Dataset

In modern cities, subways have become an important mode of transportation for urban residents, but safety issues remain a major challenge for the industry, particularly accidents that may be caused by platform boundary violations. Existing monitoring systems have shortcomings such as false positives and false negatives, unable to timely identify and warn of potential dangerous behaviors. The Subway Platform Boundary Violation Detection Image Dataset aims to address these issues by providing high-quality image data to support the training and optimization of intelligent monitoring systems, thereby enhancing the safety management capabilities of subway platforms. The dataset collection method includes obtaining images from subway platform surveillance cameras in actual operational environments, supplemented by specific scenario simulations to enrich data diversity. Quality control is ensured through multiple rounds of annotation and expert review, ensuring data consistency and accuracy. The annotation team consists of experts with backgrounds in computer vision and traffic management, with a size of 20 people. Data preprocessing uses image enhancement techniques, such as resolution adjustment and color balance, and is ultimately stored in JPG format. The structured organization facilitates inspection and use.