Stereo matching using a neural network

A polynomial is fitted to find a smooth continuous intensity function in a window and the first-order intensity derivatives are estimated. A neural network is then used to implement the matching procedure under the epipolar, photometric and smoothness constraints, using the estimated first-order derivatives. Owing to the dense intensity derivatives, a dense array of disparities is generated with only a few iterations. The method does not require surface interpolation. Computer simulations to demonstrate the efficacy of the method are presented.<<ETX>>

[1]  Parvati Dev,et al.  Perception of Depth Surfaces in Random-Dot Stereograms: A Neural Model , 1975, Int. J. Man Mach. Stud..

[2]  K. Ramesh Babu,et al.  Linear Feature Extraction and Description , 1979, IJCAI.

[3]  B. Julesz.  Binocular depth perception of computer-generated patterns , 1960 .

[4]  D Marr,et al.  Theory of edge detection , 1979, Proceedings of the Royal Society of London. Series B. Biological Sciences.

[5]  Stephen T. Barnard,et al.  A Stochastic Approach to Stereo Vision , 1986, AAAI.

[6]  John E. W. Mayhew,et al.  Psychophysical and Computational Studies Towards a Theory of Human Stereopsis , 1981, Artif. Intell..

[7]  Gene Gindi,et al.  OPTICAL NEUROCOMPUTER FOR IMPLEMENTATION OF THE MARR-POGGIO STEREO ALGORITHM. , 1987 .

[8]  N M Grzywacx,et al.  Motion correspondence and analog networks , 1987 .

[9]  Ramakant Nevatia,et al.  Segment-based stereo matching , 1985, Comput. Vis. Graph. Image Process..

[10]  Robert M. Haralick,et al.  Digital Step Edges from Zero Crossing of Second Directional Derivatives , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.