16 6 2012 12 (JKONI 16(6): 910-919, Dec. 2012) 임베디드프로세서를이용한고정익무인항공기영상기반목표물탐지및추적 김정호 *, 정재원 *, 한동인 *, 허진우 *, 조겸래 *, 이대우 * Jeong-Ho Kim *, Jae-Won Jeong *, Dong-In Han *, Jin-Woo Heo *, Kyeom-Rae Cho *, and Dae-Woo Lee * 요약,. ARM.,.,., 4. Abstract In this paper, we described development of on-board image processing system and its process and verified its performance through flight experiment. The image processing board has single ARM(Advanced Risk Machine) processor. We performed Embedded Linux Porting. Algorithm to be applied for object tracking is color-based image processing algorithm, it can be designed to track the object that has specific color on ground in real-time. To verify performance of the on-board image processing system, we performed flight test using the PNUAV, UAV developed by LAB. Also, we performed optimization of the image processing algorithm and kernel to improve real-time performance. Finally we confirmed that proposed system can track the blue-color object within four pixels error range consistently in the experiment. Key words : Embedded Processor, On-board Image Processing, UAV, Color-based Object Tracking I. 서론, * (Dept. of Aerospace Eng., Pusan National University) 1 (First Author) : (051-510-3036, kimsmap@pusan.ac.kr) (Corresponding Author) : (Dea-Woo Lee, +82-51-510-2329, baenggi@pusan.ac.kr) : 2012 11 6 ( ) : 2012 11 8 ( : 2012 12 21 ) : 2012 12 30
;,,,,, 911,,,., [1]. (GCS : Ground Control System) GCS. GCS, GCS,, [2]. [3].,. PC. ARM. 그림 1. 온보드영상처리시스템개요도 Fig. 1. The schematic diagram of Onboard image processing system 1. RGB. SD, GCS. 2-2 온보드영상처리시스템구성 2 Android S3C6410. Ⅱ. 온보드영상처리시스템개발 2-1 온보드영상처리시스템개요 그림 2. S3C6410 개발보드 Fig. 2. S3C6410 development board
912 16 6 2012 12. S3C6410 ARM11 ARM, (MFC : Multi Function Codec) [4]., 3. 표 1. S3C6410 개발보드사양 Table 1. The specifications of S3C6410 Target board 그림 3. 교차개발환경구성 Fig. 3. The structure of cross development environment (Console), OpenCV..,. OpenCV C [5][6], API V4L(Video For Linux) [7]. 2-3 온보드영상추적시스템개발절차 Host PC, Host PC, tftp, NFS(Networking File System). Host PC Ubuntu Linux 10.04(LTS), 2.6.28.6., Host PC, GCC. Host PC Host NFS. (Bootloader), FLASH ROM (0x00000000 ~ 0x0003FFFF, 256KB).
;,,,,, 913., GUI make xconfig. API V4L(Video For Linux), zimage,.. C OpenCV. LCD (Frame buffer), (Frame buffer driver). /dev /dev/fbx, LCD. LCD /dev/fb0 [8]. RGB Red 5, Green 6, Blue 5 16 (16bpp) RGB 24 16 Red Blue 3, Green 2. mmap() ㆍ.. CPU 4 (Console). 그림 4. 영상처리어플리케이션실행화면 Fig. 4. Running of the image processing application 5 CCD.., OS. 5. Fig. 5. Inner test of the image-based tracking application Ⅲ. 영상처리시스템 3-1 영상기반목표물탐지및추적알고리즘
914 16 6 2012 12., RGB. 6. 그림 7. 영상좌표계 Fig. 7. Image geometric,, 그림 6. 영상처리시스템프로세스 Fig. 6. The process of the image processing system. 8. YUV(YCbCr), YUV RGB. RGB,, DAC(Digital Analogue Converter). H.264 SD. V4L, OpenCV. 7,. 3 (1). 그림 8. 파란색물체탐지및추적알고리즘프로세스 Fig. 8. The process of detection and tracking algorithm for blue-color YUV RGB, RGB RGB 3 Red, Green, Blue 1,,,. RGB..
;,,,,, 915, 320x240(pixels) Adaptive ROI(Region of Interest). S3C6410 PP(Post Processing) RGB,. RGB.. Threshold. Adaptive ROI 10 139.61ms Adaptive ROI 307.09ms 2. 3-2 Adaptive ROI 설정 (Region of Interest), Adaptive ROI. 그림 10. Adaptive ROI 적용후평균처리시간비교 Fig. 10. Comparison of mean processing time after applying Adaptive ROI Ⅳ. 실험및결과 4-1 지상실험을통한시스템성능확인 그림 9. Adaptive ROI Fig. 9. Adaptive ROI Adaptive ROI 9,,,,,. ROI ROI 10, Threshold. Threshold ROI ROI 11. RCA,.
916 16 6 2012 12 그림 11. 무인항공기에탑재한모습 Fig. 11. PNUAV with on-board. 146.3ms..,,.. 21 320x240,.. 12,. 그림 12. 이동물체에대한탐지결과 Fig. 12. Result of detection on moving object 4-2 비행실험을통한시스템성능및결과.,, 100, SD 320x240 CPU PC(Intel Core i5 CPU 750, 2.67GHz, 4GB RAM). 13. 14 PC / /. 15., 2.95, 3.85, 3.95 4, 2.
;,,,,, 917 표 2. 비행실험결과및데스크탑 PC 와의비교 Table 2. The result of flight test and comparison between flight test and desktop PC 그림 13. 온보드영상처리수행결과 Fig. 13. The result of performance on onboard image processing Comparative on trajectory of center point(test2) 200 150 Ⅴ. 결론 Height Distance error(pixel) 100 50 Onboard Image Processing system PC(Reference) 0 0 50 100 150 200 250 300 Width 10 8 6 4 2 0-2 -4-6 -8 그림 14. 중심점위치비교 Fig. 14. Comparison of center point mean error -10 0 10 20 30 40 50 60 70 80 90 100 Frame 그림 15. 중심점거리오차 Fig. 15. Center point error, ARM,,.., 320x240(pixels) 138ms. PC Threshold 4.,.
918 16 6 2012 12 감사의글 (2 ) Reference [1] Jincheol Ha, Eric N. Johnson, and Allen Tannenbaum, Real-time Visual Tracking Using Geometric Active Contours for the Navigation and Control of UAVs, Proceedings of the 2007 American Control Conference, New York, U.S.A, 7, 2007 [2] J.H.Choi, S.H.Yim, H.C.Bang, Research for real time of On-board Image Processor, 2010 Conference of KSAS, 2010.11. pp.502~505 [3] Y.I.Kim, S.Y.Nho, Detection Method of Image Objection for Embedded System, Journal of Institute of Control, Robotics, and System, Vol.15, No.4, 2009.4, pp.422~423 [4] JK Electonics, Real 6410 Linux Development Manual [5] S.J.Lee, D.H.Won, S.C.Yoon, M.Y.Rhu, J.S.Cho, S.K.Sung, Y.J.Lee, Embodiment of Real Time Image Process System Using Embedded Processor, 2011 Conference of KSAS, 2011.11, pp.1026~1030 [6] Zheng Guo-rong, Xiong Chang-zhen, Zhang Yan, A Method of Embedded Video Surveillance based on OpenCV, E-Business and E-Government (ICEE), 2011 International Conference on, pp.1~4 [7] Paul. B. Matteucci, Member, IEEE, Philip Byrnes-Preston, Spencer C. Chen, Member, IEEE, Nigel H. Lovell, Fellow, IEEE and Gregg J. Suaning, Member, IEEE, ARM-Based Visual Processing System for Prosthetic Vision, 33rd Annual International Conference of the IEEE EMBS Boston, U.S.A, 2011 [8] J.H.An, Developmental Algorithm of Color Image Using PAX255 ARM Processor, Dept. of Computer, Dankook University, Master Thesis 김정호 ( 金政昊 ) 정재원 ( 鄭載元 ) 한동인 ( 韓東仁 ) 2008 8 : ( ) 2010 8 : ( ) 2010 9 ~ : : 2010 2 : ( ) 2010 8 : ( ) : 2010 8 : ( ) 2012 2 : ( ) 2012 3 ~ : : 허진우 ( 許鎭宇 ) 2011 8 : ( ) 2011 9 ~ : :,
;,,,,, 919 조겸래 ( 趙謙來 ) 1977 2 : ( ) 1979 2 : ( ) 1980 ( ) 1986 ( ) ( ) 1986 4 ~ : :,, 이대우 ( 李大雨 ) 1993 2 : ( ) 1997 2 : ( ) 2001 8 : ( ) 2003 3 ~ : :,,