Volume No. :   9

Issue No. :  3

Year :  2018

ISSN Print :  0976-2973

ISSN Online :  2321-581X


Allready Registrered
Click to Login

Systolic Array Based Motion Detection Architecture using Modified Reichardt Model

Address:   Abrar Ul Haq1, Sreerama Reddy G. M.2, Cyril Prasanna Raj P.3
1Assistant Professor, Dept. of Electrical and Electronics, School of Engineering, Presidency University, Itgalpur,
Rajanakunte, Yelahanka, Bengaluru-560064, Karnataka, INDIA
2Prof and Head, ECE Dept., C. Byregowda Institute of Tech, Kolar-563101
3Professor, Department of ECE, MSEC, Bangalore-562157, INDIA
*Corresponding Author
DOI No: 10.5958/2321-581X.2018.00036.3

Micro air vehicles or MAVs play a vital role in defense and domestic applications. Auto piloting of MAVs is always a challenge to hardware design engineers. Vision based autopilot aids the control strategy of MAVs for reliable flight operation. One of the important functions of vision based control logic is the motion detection of objects. Bioinspired motion detection algorithm based on Reichardt model is one approach for detection for vertical and horizontal motion of objects. In this paper, a systolic array based architecture to process 256 x 256 image data using modified Reichardt models are designed and implemented. RTL architecture has been designed using modified Reichardt correlator that can compute the motion direction in both vertical and horizontal directions. Functional simulation of motion detector has been carried out and it is observed that design takes 18 clock cycles to detect the single motion. RTL design for motion detector for 256 x 256 image sequence has been prototyped on FPGA. The design is found to operate at frequency of 322 MHz and occupies less than 300 LUTs consuming power of less than 100mW. The design is targeted and implemented on Virtex-5 FPGA optimizing area, timing and power constraints.
Reichardt model, systolic array architecture, FPGA, motion detection, MAV.
Abrar Ul Haq, Sreerama Reddy G. M., Cyril Prasanna Raj P. Systolic Array Based Motion Detection Architecture using Modified Reichardt Model. Research J. Engineering and Tech. 2018;9(3): 269-277.
[View HTML]     

Visitor's No. :   104260