MULTI-TAG OBJECT DETECTION METHOD AND SYSTEM BASED ON TARGET DETECTION NETWORK, AND APPARATUSES

摘要

Disclosed are a multi-tag object detection method and system based on a target detection network, and apparatuses, which belong to the field of computer vision, and aim to solve the problem of being unable to precisely detect a multi-tag object in real time due to the relatively low precision of existing target detection technology. The method comprises: acquiring an image of an object to be detected as an input image; using a trained target detection network to acquire, according to the input image, the category of said object and the center coordinates of said object, and the length and width of a rectangular detection frame; and outputting the category of said object and the center coordinates of said object, and the length and width of the rectangular detection frame. The method realizes real-time precise detection of different categories of targets, increases the speed and precision of detection, and can fully process the conditions of mutual overlapping and shielding during target detection.