OBJECT RECOGNITION WITH SIFT AND MI-SIFT METHODS

Yıl 2015, Cilt: 3 Sayı: 1, 15 - 25, 26.02.2015

Fırat Hardalaç , Abdullah Orman , Berkan Ural , Ali Eren

Öz

This study, which has been commonly used in object recognition area for 10 years, is based on the recognition of special point-based objects and there are differences in some areas such as luminance, anguler and resistance from cyclic and dimensional changes. Generally, scalar invariant feature transformation (scale invariant feature transform- SIFT) method and independent feature transformation (mirror reflection invariant feature- MIFT) method which is a weak point of SIFT method, also is created against the mirroring effect to gain resistance are investigated in this experiment. In this work, 11 sample trials with images which were in different cases were performed one by one to analyze these methods actually and this could give us opportunity to compare the results of the methods respectively. The aim of this work was understanding the performance capability of object recognition methods that were SIFT and MIFT in time domain and space domain, also determining the speed and result acquisition time values of these methods represented us to observe the advantages and disadvantages of SIFT and MIFT.

Anahtar Kelimeler

Object recognition, keypoint descriptors, scale invariant feature transform, SIFT, mirror reflection invariant feature transform, MIFT

Kaynakça

• [1] Lowe, D.G., Object Recognition from Local ScaleInvariant Features. Proc. of the International Conference on Computer Vision, (2): 1150-1157, 1999.
• [2] Mikolajczyk, K. Schmid, C., “A performance evaluation of local descriptors". Dept. of Eng., Oxford Univ., 78-82 2005.
• [3] Young, R. (1987). "The Gaussian derivative model for spatial vision: I. Retinal mechanisms". Spatial Vision2 (4): 273–293(21)
• [4] Davidson M. W., Abramowitz M., "Molecular Expressions Microscopy Primer: Digital Image Processing – Difference of Gaussians Edge Enhancement Algorithm", Olympus America Inc., and Florida State University.
• [5] Krystian M., Cordelia S. "A performance evaluation of local descriptors", IEEE Transactions on Pattern Analysis and Machine Intelligence, 10, 27, pp 1615-1630, 2005.
• [6] Zhao W., Ngo C., Flip-Invariant SIFT for Copy and Object Detection.IEEE Transactions on Imageprocessing, 22: 3, 2013.
• [7] Eski S., Recognition of Image Processing With Brand and Type Of Vehicle. Istanbul Technical University, 2008.
• [8] X. Guo, X. Cao, MIFT: A framework for featuredescriptors to be mirror reection invariant" Image and Vision Computing, 30546-556 (2012).
• [9] The Invariant Relations of 3D to 2D Projection of Point Sets, Journal of Pattern Recognition Research (JPRR), Vol. 3, No 1, 2008.
• [10] "The Anatomy of the SIFT Method" in Image Processing On Line, a detailed study of every step of the algorithm with an open source implementation and a web demo to try different parameters
• [11] Matthew J. McMahon, Orin S. Packer, and Dennis M. Dacey (April 14, 2004). "The Classical Receptive Field Surround of Primate Parasol Ganglion Cells Is Mediated Primarily by a Non-Gabaergic Pathway", Journal of Neuroscience.
• [12] Crowley, J, Riff O. “Fast computation of scale normalised Gaussian receptive fields”, Proc. Scale-Space'03, Isle of Skye, Scotland, Springer Lecture Notes in Computer Science, volume 2695, 2003.
• [13] Lowe, D. G. (2004). "Distinctive image features from scale-invariant keypoints". International Journal of Computer Vision 60 (2): 91–110.
• [14] Lindeberg T. (1994). "Scale-space theory: A basic tool for analysing structures at different scales". Journal of Applied Statistics (Supplement on Advances in Applied Statistics: Statistics and Images: 21 (2). pp. 224–270
• [15] Kim S., Yoon K., Kweon I., "Object Recognition Using a Generalized Robust Invariant Feature and Gestalt’s Law of Proximity and Similarity", Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'06), 2006
• [16] ezSIFT: an easy-to-use standalone SIFT implementation in C/C++. A self-contained open-source SIFT implementation which does not require other libraries.
• [17] Lowe, D. G., “Distinctive Image Features from ScaleInvariant Keypoints”, International Journal of Computer Vision, 60, 2, pp. 91-110, 2004.
• [18] Liao K., Liu G.: An improvement to the SIFT descriptor for image representation and matching,10.1016/j.patrec.2013.03.021
• [19] Soon T., Basari H., “Pattern Detection in Digital Image” 2010
• [20] Ma R., Chen J., Su Z., “MI-SIFT: Mirror and inversion invariant generalization for SIFT descriptor,” in Proc. Int. Conf. Image Video Retr., 2010, pp. 228–236.