Cellular processor array based UAV safety system

Celluular pro ocessoor arrayy based UAV V safety systtem *† Ákos Zarándy Z , Taamás Zsedrovvits†, Zoltán Nagy N *†, Andráás Kiss*†, Péteer Szolgay*†, Tamás Roskaa*† *

Computer and Automation Research Institute of the Hunngarian Academ my of Sciencess (MTA-SZTA AKI) Budapest,, Hungary † P Pázmány Péterr Catholic Univversity, The Faaculty of Inform mation Technology, Budapestt, Hungary.

Abstract — Embedded senssor-processor system is beiing A d developed for on-board UAV V (Unmanned d Aerial Vehiccle) s safety applicatioons. The role of the device is to t detect intrud der a airplanes which h are on or cllose to collision n course. Due to w weight, power, size, s and cost reequirements, th he visual approaach leeads to feasible solution only y. In our design, 5 cameras are a a applied to colleect visual data from a large field f of view. The T im mage flows arre processed by b 3 differentt virtual cellular p processor arrayys, which are im mplemented in FPGA. F

I.

Image ca apture Full frame preprocessing Gray sc cale processor Binarry processor

INTTRODUCTION

UAV safetyy system devellopments are critical c nowadaays too enable thesee autonomous flying objectss to conquer the t s sky. One of thee desperately needed n safety equipments whiich is still missingg for small UA AVs is the colllision avoidannce s system. In our demonstration n, the already operational daata a acquisition, proocessing, and arrchiving parts are a shown. The system is constructed d of 5 pieces off miniature globbal s shutter cameraas with wide VGA (WV VGA, 720x4880) r resolution (Figuure 1.). The unccompressed ouutput data stream ms o the synchroonized camerass are connecteed to the FPG of GA thhrough paralleel busses. Threee cellular proccessor arrays are a m mapped to the FPGA F (Figure 2.). The first is i responsible for f f frame proccessing. It imp full plements adapttive thresholdiing a identificatiion of the locaations of the suuspicious objects and o the entire image. on i The seecond and the third processoors w work on the sellected 128x128 8 sized window ws. They proviide thhe necessary grayscale g and binary b operatioons for detectiing thhe intruder airccrafts. The raw w flight data is saved to an SS SD f ground baseed test and verification purposses. for

Figure 1. The image capturing, processing, and storing s hardware comp ponents

978-1-4673-0289-0/12/$31.00 ©2012 IEEE

M Memory co ontroller

DRA AM

Microblaze processor

Figgure 2. Block diaagram of the image processing archiitecture

II.

DESCR RIPTION OF THE E DEMONSTRAT TOR

Durring the demonnstration, the first fi version off the device (Figuree 3.) will be shown. It will bee able operate either e from images collected throough its opticaal input, from previously acquireed videos, or onn-line synthesizzed scenes.

Figure 3.. Early version of the collision avoidance sensor proccessor system

ACKNOWLEDGMENT The ONR Grants (N62909-11-1-7039, N62909-10-17081) is greatly acknowledged. REFERENCES [1] [2]

[3]

[4] [5]

[6]

[7]

[8]

http://en.wikipedia.org/wiki/Aircraft_collision_avoidance_systemsAF RL Sense and avoid Debadeepta Dey, Christopher Geyer, Sanjiv Singh and Matt Digioia “Passive, Long-Range Detection of Aircraft: Towards a Field Deployable Sense and Avoid System” Field and Service Robotics, Springer Tracts in Advanced Robotics, 2010, Volume 62/2010, 113123, DOI: 10.1007/978-3-642-13408-1_11 John Lai, Luis Mejias, Jason J. Ford, “Airborne vision-based collision-detection system”, Journal of Field Robotics, Volume 28, Issue 2, pages 137–157, March/April 2011 ICAO Doc 4444 (Air Traffic Management) T. Zsedrovits, Á. Zarándy, B. Vanek, T. Péni, J. Bokor, T. Roska, „Collision avoidance for UAV using visual detection” ISCAS-2011, Rio de Janeiro, Brasil Z. Nagy, A. Kiss, Á. Zarándy, B. Vanek, T. Péni, J. Bokor, T. Roska “Volume and power optimized high-performance system for UAV collision avoidance” ISCAS-2012, Seoul, Korea Z. Voroshazi, A. Kiss, Z. Nagy, and P. Szolgay, “Implementation of embedded emulated-digital CNN-UM global analogic programming unit on FPGA and its application,” International Journal of Circuit Theory and Applications, vol. 36, no. 5-6, pp. 589–603, 2008. “Xilinx products homepage,” [Online] http://www.xilinx.com, 2011.