Three-dimensional holographic imaging security radar image foreign object detection method

Abstract

The invention relates to the technical field of holographic imaging security inspection, in particular to a three-dimensional holographic imaging security inspection radar, and specifically provides a method for detecting foreign matter in a three-dimensional holographic imaging security inspection radar image. The foreign matter detection method includes the following steps: acquiring 3D radar image target detection data; inputting the 3D radar image target detection data into a training convergent anchor-free neural network model, and obtaining the 3D radar output from the anchor-free neural network model Image foreign object detection results. The present invention can greatly reduce the number of prediction frames extracted by the neural network model, reduce the computational complexity of the model, and make the final predicted detection frame have good scalability, which helps to improve the recognition effect of objects with large scale changes; At the same time, through the multi-level 3D radar image feature extraction method, the neural network model can also obtain more sufficient semantic information, which is conducive to the detection and recognition of objects in the image scene.

Description

Technical field: [0001] The invention relates to the technical field of holographic imaging security inspection, in particular to a three-dimensional holographic imaging security inspection radar, and specifically provides a method for detecting foreign matter in a three-dimensional holographic imaging security inspection radar image. Background technique: [0002] With the development of society and economy, airports, railway stations, subways, bus stations and other public transportation hubs have ushered in more and more passenger flows and logistics. The security inspection level of these areas must eliminate potential safety hazards to ensure public safety and meet High detection efficiency with fast pass. Traditional foreign object detection methods mainly use security inspection technologies such as metal detection and X-ray imaging, which have gradually failed to meet the needs of security inspection. Developed countries in Europe and the United States have graduall...

AI Technical Summary

Problems solved by technology

[0004] Three-dimensional holographic imaging security radar is a newly developed security detection technology. At present, in terms of foreign object detection in three-dimensional holographic imaging security radar images at home and abroad, it is mainly based on conventional optical image target detection methods; due to radar images affected by coherent speckle noise and observation Due to the influence of viewing angle, the applicability of conventional training, verification and testing methods for object detection and recognition in optical images is limited to a certain extent

Images