Automatic Target Recognition for CT-based Airport Screening System

Abstract

To construct an automatic target recognition CT-based screening system for airport, this paper introduces a method by combining two hot technology, CT Image Processing and Machine Learning together. With the grey-scale features and the Histogram of Oriented Gradient (HOG) features extracted from the CT images, we can train different classifier (SVM, KNN) to recognize the targets (saline, rubber, clay) we want. By comparing the recall rate and precision rate of each classifier, we may find the best classifier for this project.