MoR: Moving Object Recognition in Video Using Hybrid Approach

Abstract

Computer vision and intelligent video surveillance system are more interesting topics in the field of moving object recognition. The security systems which are effective will have intelligent video surveillance as an integral part of it. The major worldwide concern is security since any criminal activities occurred across the world. Monitoring such events currently rely on man power and technology; however, in order to avoid human errors by using advanced automatic monitoring technology that can be affected by different reasons. To overcome these shortfalls, the intelligent surveillance system is developed for monitoring multiple moving object recognitions. Object recognition remains challenging due to illumination shadows, changes, and occlusion. All these make it necessary to develop robust approaches. Gabor–PCA approach and distance similarity technique are proposed for multiple moving object recognitions such as a human, vehicle, etc. The proposed approach achieves good recognition performance under complex situations with mostly datasets. Thus, the system MoR provides a simple, efficient, and rapid solution to the problem of recognizing multiple moving objects.