Görsel Odometride SIFT, SURF, FAST, STAR ve ORB özellik algılama algoritmalarının Performans ve Takas Değerlendirmesi

Year 2020, Ejosat Special Issue 2020 (ICCEES), 455 - 460, 05.10.2020

Abdullah Yusefı , Akif Durdu , Cemil Sungur

https://doi.org/10.31590/ejosat.819735

Cited By: 1

Abstract

Son yıllarda görsel odometri alanında, robotik ve otomotiv teknolojisinde görsel tabanlı ölçüm gibi pratik süreçlerin geliştirilmesine yol açan çok sayıda araştırma ve çalışma yapılmıştır. Doğrudan yöntemler, özellik tabanlı yöntemler ve hibrit yöntemler, görsel odometri problemlerinin çözümünde üç yaygın yaklaşımdır ve öznitelik tabanlı yaklaşım hızlarının daha yüksek olduğu genel inancı göz önüne alındığında, bu yaklaşım son yıllarda memnuniyetle karşılanmaktadır. Bu nedenle, bu çalışmada, kamera dönüşü ve çevirisindeki değişiklikleri tahmin edebilen değişmez özellikler kullanılarak iki boyutlu sıralı görüntü setlerinin dönüşüm matrisini hesaplamak için bir girişimde bulunulmuştur. Algoritmada, anahtar noktaların belirlenmesi ve aykırı değerlerin kaldırılması için iki adım, sırasıyla beş farklı yerel özellik algılama algoritması (SURF, SIFT, FAST, STAR, ORB) ve RANdom SAmple Consensus algoritması (RANSAC) kullanılarak gerçekleştirilir. Ek olarak, her birinin etkisi, içsel parametreleri ve dinamik gürültü, dönüşüm matrisinin doğruluğu üzerinde değerlendirilir ve rotasyonel MSE ve hesaplama çalışma süresi açısından analiz edilir.

Keywords

Görsel Odometri, Görüntü İşleme, Değişmeyen Özellikler, Yerel Özellik Algılama, Anahtar Noktalar, RANSAC, Dönüşüm Matrisi

Performance and Trade-off Evaluation of SIFT, SURF, FAST, STAR and ORB feature detection algorithms in Visual Odometry

Abstract

In recent years there has been a great deal of research and study in the field of visual odometry, which has led to the development of practical processes such as visual based measurement in robotics and automotive technology. Direct methods, feature-based methods and hybrid methods are three common approaches in solving visual odometry problems and given the general belief that feature-based approach speeds are higher, this approach has been welcomed in recent years. Therefore, an attempt has been made in the present study to calculate the transformation matrix of two-dimensional sequential image sets using invariant features that can estimate the changes in camera rotation and translation. In the algorithm, two-steps of identifying keypoints and removing outliers are performed using five different local feature detection algorithms (SURF, SIFT, FAST, STAR, ORB) and RANdom SAmple Consensus algorithm (RANSAC), respectively. In addition, the impact of each of them, their intrinsic parameters and dynamic noise on the accuracy of the transformation matrix are evaluated and analyzed in terms of rotational MSE and computational runtime.

Keywords

Visual Odometry, Image Processing, Invariant Features, Local Feature Detection, Keypoints, RANSAC, Transformation Matrix

References

• Nistér, David, Oleg Naroditsky, and James Bergen. "Visual odometry." Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004.. Vol. 1. Ieee, 2004.
• Scaramuzza, Davide, and Friedrich Fraundorfer. "Tutorial: visual odometry." IEEE Robotics and Automation Magazine 18.4 (2011): 80-92.
• Fraundorfer, Friedrich, and Davide Scaramuzza. "Visual odometry: Part ii: Matching, robustness, optimization, and applications." IEEE Robotics & Automation Magazine 19.2 (2012): 78-90.
• Civera, Javier, et al. "1-point RANSAC for EKF-based structure from motion." 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2009.
• Scaramuzza, Davide. "1-point-ransac structure from motion for vehicle-mounted cameras by exploiting non-holonomic constraints." International journal of computer vision 95.1 (2011): 74-85.
• Rosten, Edward, and Tom Drummond. "Machine learning for high-speed corner detection." European conference on computer vision. Springer, Berlin, Heidelberg, 2006.
• Rosten, Edward, and Tom Drummond. "Fusing points and lines for high performance tracking." Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1. Vol. 2. Ieee, 2005.
• Konolige, Kurt, Motilal Agrawal, and Joan Sola. "Large-scale visual odometry for rough terrain." Robotics research. Springer, Berlin, Heidelberg, 2010. 201-212.
• Harris, Christopher G., and Mike Stephens. "A combined corner and edge detector." Alvey vision conference. Vol. 15. No. 50. 1988.
• Wei, Lijun, et al. "GPS and stereovision-based visual odometry: Application to urban scene mapping and intelligent vehicle localization." International Journal of Vehicular Technology 2011 (2011).
• Lowe, David G. "Distinctive image features from scale-invariant keypoints." International journal of computer vision 60.2 (2004): 91-110.
• Bay, Herbert, Tinne Tuytelaars, and Luc Van Gool. "Surf: Speeded up robust features." European conference on computer vision. Springer, Berlin, Heidelberg, 2006.
• Fischler, Martin A., and Robert C. Bolles. "Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography." Communications of the ACM 24.6 (1981): 381-395.
• Lindeberg, Tony. "Feature detection with automatic scale selection." International journal of computer vision 30.2 (1998): 79-116.
• Agrawal, Motilal, Kurt Konolige, and Morten Rufus Blas. "Censure: Center surround extremas for realtime feature detection and matching." European Conference on Computer Vision. Springer, Berlin, Heidelberg, 2008.
• Poddar, Shashi, Rahul Kottath, and Vinod Karar. "Evolution of visual odometry techniques." arXiv preprint arXiv:1804.11142 (2018).
• E. Rublee, et al. "ORB: An efficient alternative to SIFT or SURF." 2011 International conference on computer vision. Ieee, 2011.
• Klette, Reinhard. Concise computer vision. Springer, London, 2014.
• Mur-Artal, Raul, and Juan D. Tardós. "Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras." IEEE Transactions on Robotics 33.5 (2017): 1255-1262.
• Corke, Peter, Dennis Strelow, and Sanjiv Singh. "Omnidirectional visual odometry for a planetary rover." 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)(IEEE Cat. No. 04CH37566). Vol. 4. IEEE, 2004.
• Scaramuzza, Davide. "Performance evaluation of 1‐point‐RANSAC visual odometry." Journal of Field Robotics 28.5 (2011): 792-811.
• Nistér, David. "An efficient solution to the five-point relative pose problem." IEEE transactions on pattern analysis and machine intelligence 26.6 (2004): 756-770.
• Tardif, Jean-Philippe, Yanis Pavlidis, and Kostas Daniilidis. "Monocular visual odometry in urban environments using an omnidirectional camera." 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2008.
• Govender, Natasha. "Evaluation of feature detection algorithms for structure from motion." (2009).
• Leutenegger, Stefan, Margarita Chli, and Roland Y. Siegwart. "BRISK: Binary robust invariant scalable keypoints." 2011 International conference on computer vision. Ieee, 2011.
• Chien, Hsiang-Jen, et al. "When to use what feature? SIFT, SURF, ORB, or A-KAZE features for monocular visual odometry." 2016 International Conference on Image and Vision Computing New Zealand (IVCNZ). IEEE, 2016.
• Bauer, Johannes, Niko Sünderhauf, and Peter Protzel. "Comparing several implementations of two recently published feature detectors." IFAC Proceedings Volumes 40.15 (2007): 143-148.