An efficient method for the exact computation of Legendre moments of binary images
Zhifang Dong, Hui Zhang, Huazhong Shu · 2009
Legendre moments have been widely used as feature descriptors in image analysis. This paper presents a new approach for the exact computation of Legendre moments of binary images. Using a Green's formula, the double integral is first transformed into a simple integration along the boundary which is represented by a polygon. We then propose an efficient method for computing the line moments. Comparison with some known algorithms shows that the proposed algorithm is more accurate, and is independent of the slope, so that it is stable in numerical computation.