Image pre-classification method of airport security inspection contraband automatic identification system

Abstract

The invention provides an image pre-classification method of an airport security inspection contraband automatic identification system. The method comprises the following steps: S1; acquiring front illumination and side illumination of an X-ray machine image; finishing RGB and HSI conversion of an input image through side illumination and front illumination, using color segmentation to judge whether a black area exists in the image or not through a black threshold value Vth, and when the area is larger than 3 * 3 pixel the front illumination and the side illumination both have unknown areas, an alarm can be given to remind security personnel of unpacking inspection; S2, filtering and denoising the frontal images and the side images processed in the step S1, and extracting HOG features andLBP features; s3, constructing a comprehensive feature vector X (HOG _ V, LBP _ V, HOG _ S and LBP _ S) in a serial cascade mode; and S4, pre-classifying the comprehensive feature vectors X (HOG _ V,LBP _ V, HOG _ S and LBP _ S) of the frontal illumination and the side illumination of the prohibited goods by adopting an SVM (Support Vector Machine) classifier, and immediately sending true to an alarm to trigger alarming when the prohibited goods exist. According to the invention, the front-illuminated X-ray image and the side-illuminated X-ray image are fused, the combined feature vector is constructed, coarse classification of prohibited goods is achieved, and the overall robustness and real-time performance of the classification algorithm are improved.

Description

technical field [0001] The invention relates to the field of airport security inspection, in particular to an image pre-classification method for an auxiliary recognition system for contraband in airport security inspection luggage. Background technique [0002] With the development of my country's economy, the civil aviation industry has entered the fast lane of development, and the construction of airports as supporting facilities for the civil aviation industry has also entered a stage of rapid development. Especially in recent years, the passenger throughput of the airport has increased year by year, and the security pressure demand of the airport security inspection system has increased year by year. Whether it can effectively protect the flight safety of aircraft and comprehensively improve the safety precautions of the airport has become an urgent problem to be solved. As a transportation hub, the airport has a large flow of people, and the level of security protecti...

AI Technical Summary

Problems solved by technology

However, at present, the X-ray machine at the airport is still in the traditional monitoring stage of the map judge. In the traditional mode, the map judge needs to keep an eye on and monitor the video screen of the X-ray machine with high intensity, and the work intensity is very high.

Although map judges generally "shift half an hour", with the continuous emergence of new types of dangerous goods, which far exceed the limit of human beings, it poses a severe challenge to manual map judgment methods relying on memory and experience. Misjudgments or even missed judgments are often the case. occur

Images