Automatic object searching system based on Real Time SIFT Algorithm

Abstract

The field of object searching and recognition systems is an increasingly popular area of focus for industry and research labs alike. Object searching mechanism is a challenging task. The proposed system makes use of computer vision techniques. Real time object search poses many difficulties in recognizing objects. These difficulties can be overcome by Triangular Matrix Search Algorithm and Real Time SIFT Algorithm, which is an improved version of scale invariant feature transform(SIFT) for real time situations. This paper presents the results of experimental studies on object searching systems performance. For this an integrated hardware platform has developed, which consists of PTZ IP-camera and embedded system based wheeled robot.