Wavelets/multiwavelets in stereo correspondence estimation: a comparative study
Asim Bhatti, Saeid Nahavandi · Deakin Research Online (Deakin University) · 2009
Stereo correspondence estimation in one of the most active research areas in the field of computer vision and number of techniques has been established possessing both advantages and disadvantages. Among the reported techniques, multiresolution analysis based stereo correspondence estimation has gained lot of research focus in recent years, however, relatively little work has been reported in the context of wavelets/multiwavelts based multiresolution analysis. Wavelets and multiwavelets analysis is a newer way of scale space representation and considered to be as fundamental as Fourier representation and a better alternative. In this work we have tried to address some of the issues regarding the work done in this context and a new technique is devised to overcome some of the flaws that has not been addressed comprehensively. For this purpose a new multi-resolution analysis based algorithm is presented that makes use of the wavelets/multiwavelts transform modulus maxima to establish correspondences between the stereo pair of images. Furthermore, variety of wavelets and multiwavelets bases, possessing distinct properties such as orthogonality, approximation order, shape, etc., are employed, to analyse their effect on the performance of correspondence estimation. The idea is to provide knowledge base to understand and establish relationships between wavelets and multiwavelets properties and their effect on the quality of stereo correspondence estimation. Keywords-Stereo Correspondence estimation; wavelets; multiwavelets; multiresolution analysis; stereo vision