Abstract
We propose an approach to image comparison that accounts for deformations and lighting changes in the object being viewed. We address this problem by defining a Riemannian manifold of the space of all images, in which the geodesic distance between two points represents the distance between two images. In order for this manifold to capture the effects of lighting and deformation, we define a local image metric that measures deformations and intensity changes. In particular, the component of our metric that handles intensity variations is designed to penalize changes less if they are likely to be due to lighting variation. We provide some theoretical properties of the resulting geodesic distances. We also show the potential value of this new local metric, by incorporating it into an optical flow framework and then showing that this can be used for face recognition. Finally, we show that the lighting-insensitive cost for intensities that we introduce allows us to compute an approximation to geodesic distances for this component of our metric very efficiently, by working in the wavelet domain.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Aggarwal JK, Cai Q (1999) Human motion analysis: a review. Comput Vis Image Underst 73:90–102
Baker S, Matthews I (2004) Lucas-Kanade 20 years on: a unifying framework. Int J Comput Vis 56(3):221–255
Beg MF, Miller MI, Trouvé A, Younes L (2005) Computing large deformation metric mappings via geodesic flows of diffeomorphisms. Int J Comput Vis 61(2):139–157
Brox T, Bruhn A, Papenberg N, Weickert J (2004) High accuracy optical flow estimation based on a theory for time warping. In: ECCV, vol 4, pp 25–36
Chen H, Belhumeur P, Jacobs D (2000) In search of illumination invariants. In: IEEE proc comp vis and pattern recognition, vol I, pp 254–261
Cootes TF, Taylor CJ (2001) Statistical models of appearance for medical image analysis and computer vision. In: Proc. SPIE medical imaging, pp 236–248
Criminisi A, Blake A, Rother C, Shotton J, Torr PHS (2007) Efficient dense stereo with occlusions for new view-synthesis by four-state dynamic programming. Int J Comput Vis 71(1):89–110
Durrleman S, Pennec X, Trouvé A, Thompson P, Ayache N (2008, in press) Inferring brain variability from diffeomorphic deformations of currents: an integrative approach. Med Image Anal
Gopalan R, Jacobs D (2010) Comparing and combining lighting insensitive approaches for face recognition. Comput Vis Image Underst 114:135–145
Hager G, Belhumeur P (1998) Efficient region tracking with parametric models of geometry and illumination. IEEE Trans Pattern Anal Mach Intell 20(10):1125–1139
James AP (2010) Pixel-level decisions based robust face image recognition. In: Oravec M (ed) Face Recognition, chap 5. INTECH, pp 65–86
Jorstad A, Jacobs D, Trouvé A (2011) A deformation and lighting insensitive metric for face recognition based on dense correspondence. In: IEEE conference on computer vision and pattern recognition (CVPR)
Jorstad A (2012) Measuring deformations and illumination changes in images with applications to face recognition. PhD thesis, University of Maryland
Ling H, Jacobs D (2005) Deformation invariant image matching. In: IEEE international conference on computer vision, vol II, pp 1466–1473
Martinez A (2003) Recognizing expression variant faces from a single sample image per class. In: CVPR, vol 1, pp 353–358
Martinez A, Benavente R (1998) The AR face database. CVC technical report #24
Miller MI, Trouvé A, Younes L (2006) Geodesic shooting for computational anatomy. J Math Imaging Vis 24(2):209–228
Negahdaripour S (1998) Revised definition of optical flow: integration of radiometric and geometric cues for dynamic scene analysis. IEEE Trans Pattern Anal Mach Intell 20:961–979
Osadchy M, Jacobs D, Lindenbaum M (2007) Surface dependent representations for illumination insensitive image comparison. IEEE Trans Pattern Anal Mach Intell 29(1):98–111
Shirdhonkar S, Jacobs D (2008) Approximate earth movers distance in linear time. In: CVPR
Song J, Chen B, Wang W, Ren X (2008) Face recognition by fusing binary edge feature and second-order mutual information. In: IEEE conf on cybernetics and intelligent systems, pp 1046–1050
Trouvé A, Younes L (2005) Local geometry of deformable templates. SIAM J Math Anal 37(1):17–59
Trouvé A, Younes L (2005) Metamorphoses through Lie group action. Found Comput Math 5(2):173–198
Zhao S, Gao Y (2008) Significant jet point for facial image representation and recognition. In: ICIP, pp 1664–1667
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag London
About this chapter
Cite this chapter
Jacobs, D., Jorstad, A., Trouvé, A. (2013). Deformations and Lighting. In: Dickinson, S., Pizlo, Z. (eds) Shape Perception in Human and Computer Vision. Advances in Computer Vision and Pattern Recognition. Springer, London. https://doi.org/10.1007/978-1-4471-5195-1_9
Download citation
DOI: https://doi.org/10.1007/978-1-4471-5195-1_9
Publisher Name: Springer, London
Print ISBN: 978-1-4471-5194-4
Online ISBN: 978-1-4471-5195-1
eBook Packages: Computer ScienceComputer Science (R0)