Parallel Design of Background Subtraction and Template Matching Modules for Image Objects Tracking System

In recent years, many researchers have proposed intelligent systems based on the IoT (Internet of Things). Among these smart systems, one of the most common applications is intelligent surveillance system. Due to the development of the camera, some applications adopt higher resolution of images to get more accurate results. Therefore, how to process these high-resolution images in real time has become more and more important. In this paper, we design two efficient libraries to detect and track objects. Background subtraction and template matching techniques are our basic approaches which are usually applied to object detection and tracking systems. In order to process high-resolution images, we optimize these two modules by parallel technique to enhance the performance. Experimental results show that the performance of the tracking system using the proposed approach can be increased about 52%.