Author : Rumaisa Iftikhar, Samiya Ashgar, Hira Abbasi, Tahir Qadri, Sadia Abbasi and Naeem Abbas

Digital Video processing technology is maturing rapidly with applications widely spread from everyday life to hi-tech engineering. Execution of real-time video processing requires good memory performance and high computational load; therefore its implementation is frequently confined with expensive DSP kits or large workstations. FPGA has proven to be an attractive and efficient alternative for such systems because of its flexibility and the added benefit of portability. An important parameter of video processing is ability to detect moving objects from the frame of the video. This object detection is further enhanced to gesture recognition which is gaining gigantic interest in various fields including image processing and computing comprising various factors as well as constraints nonetheless providing ease in numerous human affairs. In the presented paper, the field of gesture recognition has been narrowed to hand gesture recognition, in particular to palm of the hand used effectively to control the movement of the cursor. A simple but effectual system has been developed where image of a input hand is computed. The approach includes calculation of centroid of the hand where the system is invariant to the scale of the hand. The proposed model is implemented on Altera’s DE2-115 FPGA board with input feed from HD webcam C270h capturing 1280 x 720 pixels in YCbCr format. The real life video processing results illustrates that the presented procedure has effective implementation.

Keywords: Video Processing, FPGA, Controlling of cursor, Calculation of centroid

Article published in International Journal of Current  Engineering  and Technology, Vol.4,No.2 (April- 2014)