실시간 모델기반 카메라 추적을 위한 속도 지향적 SURF 알고리즘 개발
Development of Speed-oriented SURF Algorithm for Real Time Model-based Camera Tracking
4
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
SURF는 크기, 평면상의 회전, 밝기변화 등에 강인하게 특징점을 검출, 추적할 수 있다. 또한 SURF는 적분 영상(integral image)을 사용하여 영상 콘볼루션을 수행하기 때문에 다른 특징점 검출, 추적 방법에 비해 빠르다. 이러한SURF의 장점으로 인해 SURF는 특징점 추적을 기반으로 하는 다양한 응용 분야에 활용되고 있다. 그러나 SURF는카메라 추적과 같은 실시간 응용 분야에 활용될 만큼 빠르지 못하다. 그러므로 본 논문은 SURF를 카메라 추적에 적용하기 위해 SURF의 속도를 개선하기 위한 4가지 효과적인 방법(광류기반 추적 방법과의 결합, 영상 크기 축소, 순차적 콘볼루션에 의한 고속 Hessian 계산, 특징점 서술을 위한 샘플점 수의 감소)을 제안한다. 또한 속도가 개선된SURF를 기 개발된 모델기반 카메라 추적 시스템에 적용하여 성능을 검증한다. 결과로서 개선된 SURF를 사용한 모델기반 카메라 추적 시스템은 일반적인 노트북 환경에서 640×480 크기의 카메라 영상에 대해 7.7fps에서 24.6fps으로 동작속도가 향상되었다.
Abstract
Speeded Up Robust Features (SURF) is invariant to scale, in-plane rotation, and contrast change with a bias in intensities. In addition, it is faster than other state-of-the-art point feature detection/tracking algorithms due to the use of integral images for image convolutions. These advantages led researchers to attempting to use SURF for various applications. However, SURF was not as fast as being applicable to real-time applications such as camera tracking. Therefore, we carefully explore effective methods for accelerating SURF and propose four methods: cooperative integration of an optical flow method to SURF, image size reduction using pixel area relation, fast Hessian computation by an ordinal convolution method, and feature description at fewer sample points, in this paper. Then, we apply the speed-oriented SURF to a hybrid model-based camera tracking system. As a result, the model-based camera tracking system could work at 24.6 fps (originally worked at 7.7 fps) with camera images of a resolution of 640×480 pixels in an ordinary laptop computer.
참고문헌 (31)
- M. Agrawal, "CenSurE: Center surround extremas for realtime feature detection and matching", Proc. of European Conference on Computer Vision, pp.102~115 (2008)
- H. Bay, "Speeded-up robust features (SURF)", Computer Vision and Image Understanding, 110(3), pp.346~359 (2008)
- H. Bay, "SURF: Speeded up robust features", Proc. of European Conference on Computer Vision, pp.404~417 (2006)
- M. Brown, "Invariant features from interest point groups", Proc. of British Machine Vision Conference, pp.656~665 (2002)
- M. Brown, "Automatic panoramic image stitching using invariant features", International Journal of Computer Vision, 74, pp.59~77 (2007)
- J. Canny, "A computational approach to edge detection", IEEE Trans. on Pattern Analysis and Machine Intelligence, 8, pp.679~714 (1986)
- N. Cornelis, "Fast scale invariant feature detection and matching on programmable graphics hardware", Proc. of IEEE Conference on Computer Vision and Pattern Recognition Workshop, pp.1~8 (2008)
- T. Drummond, "Real-time visual tracking of complex structures", IEEE Trans. on Pattern Analysis and Machine Intelligence, 24(7), pp.932~946 (2002)
- M. Ebrahimi, "SUSurE: Speeded up surround extrema feature detector and descriptor for realtime applications", Proc. of IEEE Conference on Computer Vision and Pattern Recognition Workshop, pp.9~14 (2009)
- D. Gossow, "An evaluation of open source SURF implementations" (2010), RoboCup 2010: Robot Soccer World Cup XIV
- B. L. V. Gouet, "SAP: A robust approach to track objects in video streams with snakes and points", Proc. of British Machine Vision Conference, pp.737~746 (2004)
- B. D. Lucas, "An iterative image registration technique with an application to stereo vision", Proc. of Image Understanding Workshop, pp.121~130 (1981)
- K. Mikolajczyk, "A performance evaluation of local descriptors", IEEE Trans. on Pattern Analysis and Machine Intelligence, 27(10), pp.1615~1630 (2005)
- K. Mikolajczyk, "A comparison of affine region detectors", International Journal of Computer Vision, 65, pp.43~72 (2005)
- H. Park, "Analytic fusion of visual cues in model-based camera tracking", Proc. of International Conference on Virtual-Reality Continuum and Its Applications in Industry (2009)
- 서병국, "Automatic confidence adjustment of visual cues in model-based camera tracking", COMPUTER ANIMATION AND VIRTUAL WORLDS, 21(2), pp.69~79, John Wiley & Sons, Ltd.
- 박한훈, "Fast Detection of Robust Features by Reducing the Number of Box Filtering in SURF", IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS, E94D(3), pp.725~728, IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG
- E. Rosten, "Fusing points and lines for high performance tracking,'' in Proc. of International Conference on Computer Vision, 2005, pp", Proc. of International Conference on Computer Vision, pp.1508~1515 (2005)
- B. Ruf, "Mobile museum guide based on fast SIFT recognition", Proc. of International Workshop on Adaptive Multimedia Retrieval (2008)
- N. Snavely, "Photo tourism: Exploring photo collections in 3D", ACM Trans. on Graphics, 25(3), pp.835~846 (2006)
- J. Svab, "FPGA based speeded up robust features", Proc. of IEEE Conference on Technologies for Practical Robot Applications, pp.35~41 (2009)
- D.-N. Ta, "SURFTrac: Efficient tracking and continuous object recognition using local feature descriptors", Proc. of IEEE Conference on Computer Vision and Pattern Recognition, pp.2937~2944 (2009)
- T. B. Terriberry, "GPU accelerating speeded-up robust features", Proc. of International Symposium on 3D Data Processing, Visualization and Transmission, pp.355~362 (2008)
- L. Vacchetti, "Combining edge and texture information for real-time accurate 3D camera tracking", Proc. of International Symposium on Mixed and Augmented Reality, pp.48~57 (2004)
- P. Viola, "Rapid object detection using a boosted cascade of simple features", Proc. of IEEE Conference on Computer Vision and Pattern Recognition, pp.511~518 (2001)
- N. Zhang, "Computing optimized parallel speeded-up robust features (P-SURF) on multi-core processors", International Journal of Parallel Programming, 38(2), pp.138~158 (2010)