A Study on Optimization Methods of X-Ray Machine Recognition for Aviation Security System

Traditional X-ray machine image recognition methods for airport security system have difficulties in recognition and are prone to result in recognition errors due to the impact of placing angle, density and volume of detected objects. This paper accurately describes the image features of X-ray machine visual image, carries out SVM classification after a visual dictionary is formed and enhances the accuracy of image discrimination by means of robust acceleration. The experimental results indicate that both identification efficiency and accuracy are improved to some extent.