APPLICATION OF SUPPORT VECTOR MACHINE CLASSIFIER FORSECURITY SURVEILLANCE SYSTEM

摘要

This paper presents the application of Support VectorrnMachine classifier for security surveillance system.rnRecently, research in image processing has raised muchrninterest in the security surveillance systems community.rnWeapon detection is one of the greatest challenges facing byrnthe community recently. In order to overcome this issue,rnapplication of the popularly used Support Vector Machinernclassifier is performed to focus on the need of detectingrndangerous weapons. In this paper, we take advantage of thernclassifier to categorize images object with the hope to detectrndangerous weapons effectively. In order to validate therneffectiveness of Support Vector Machine classifier, severalrnclassifiers are used to compare the overall accuracy of thernsystem. These classifiers include Neural Network, DecisionrnTrees, Na?ve Bayes and k-Nearest Neighbor methods. Thernfinal outcome of this research clearly indicates that SupportrnVector Machine has the ability in improving thernclassification accuracy using the extracted features.