Abstract:
In our work we present generalization of well-known approach for construction of invariant feature vectors of images in computer vision applications. Basic feature of the suggested algorithm is replacement of commonly used Gaussian filter by convolution of image function with Green’s function of evolution operator, which inherits symmetries of this operator. The use of general filtration allows to obtain additional characteristics of invariant feature vectors.
Keywords:
computer vision, pattern recognition, filtering, symmetries of differential equations, Green Function.
Citation:
D. A. Vrazhnov, A. V. Shapovalov, V. V. Nikolaev, “Symmetries of differential equations in computer vision applications”, Computer Research and Modeling, 2:4 (2010), 363–376