A multi-target detection method

Abstract

This invention relates to a multi-target detection method, comprising: effectively decoupling the track texture from environmental background noise based on anisotropic filtering, significantly improving feature representation and signal-to-noise ratio in low-contrast and complex texture environments; locating and repairing fracture areas caused by partial occlusion, reflection, and other adverse conditions on the track, ensuring the connectivity and integrity of the output trajectory at the topological level; and eliminating spatial semantic ambiguity caused by monocular vision perspective effects by constructing a feasible domain mask surrounding the track, ensuring that foreign object alarms are strictly limited to physical danger zones, and greatly reducing the false alarm rate. By simultaneously extracting the railway track topology and investigating foreign object risks in the track area, the detection accuracy under complex conditions is significantly improved.

Images