X-ray contraband package detection method and device based on feature map re-empowerment

Abstract

The invention discloses an X-ray contraband package detection method and device based on feature map re-empowerment, and the method comprises the steps: building an X-ray contraband training set, sequentially obtaining an image feature map, a region-of-interest feature map and a class attention feature map, carrying out the re-empowerment of the region-of-interest feature map through the class attention feature map, and obtaining a single-dimensional feature vector; determining prediction categories of various contraband products, adjusting a feature extraction network, a meta-feature extractor and a classification regression network adopted in the next training process, and returning to execute the process of sequentially acquiring an image feature map, a region-of-interest feature map and a class attention feature map; and inputting the X-ray contraband training set into the adjusted feature extraction network again for iterative training until the number of training times is greaterthan or equal to a preset number of iterations, determining a contraband detector, and inputting the to-be-detected package into the contraband detector to obtain a contraband position and a contraband category output by the contraband detector, thereby achieving the effects of saving manpower and material resources and reducing the detection cost.

Description

technical field [0001] The invention relates to the technical field of intelligent detection, in particular to a method, device, computer equipment and storage medium for X-ray contraband package detection based on feature map reweighting. Background technique [0002] X-ray images are images obtained by X-rays penetrating the object to be detected and rendering different colors according to the different densities of the object. It is often used in security checks to identify contraband. Compared with ordinary images, X-ray images only have color and shape information, and it is more difficult to detect and identify objects. Therefore, X-ray contraband detection has always been a difficult problem to be solved. [0003] Especially in recent years, the development of the transportation industry and the fast industry has caused enormous pressure on luggage / express security checks. In order to ensure the safety of public areas, before taking a large vehicle, the items carried...

AI Technical Summary

Problems solved by technology

For safety reasons, traditionally, professional security inspectors are required to manually observe the X-ray image of each package next to the X-ray machine to confirm whether the package contains contraband, which requires a lot of manpower and material resources, and the work efficiency increases with the security personnel. The fatigue degree increases and decreases, and the cases of missed detection and false detection will increase accordingly

[0004] At present, the main method of identifying contraband in parcels is mainly to rely on manpower for security inspection, which usually requires several groups of staff to work in turn, and each group includes at least one professional security inspector and two sorting personnel, so that the traditional package detection scheme exists. The problem of low efficiency and high cost

Images