Target detection method and system based on radar and image data fusion

The invention discloses a target detection method and system based on radar and image data fusion. The method comprises the steps of object detection of an encoder network, data association of a view cone method and image feature supplementation of radar detection. The system comprises an inspection area environment sensing module, a millimeter wave signal processing module, a point cloud analysis and processing module, a deep learning computer and a display module, and the modules are in wired or wireless transmission. According to the invention, the millimeter-wave radar is used as environment sensing sensor equipment of the inspection area, so that the influence of environmental factors such as light on the security system performance is effectively avoided; by reading the radar point cloud data of the sub-region where the target is located for processing and analysis, the calculation amount caused by processing non-dynamic redundant information of the inspection region by a deep learning computer is reduc