Jacobi moments as image features

In this paper, we introduce a set of moments which is based on Jacobi polynomials. The set of Jacobi polynomials are orthogonal and this ensures minimal information redundancy between the moments. By changing the parameters /spl alpha/ and /spl beta/, it is shown that the moments are able to extract both global and local features. This is unseen in moments such as geometric, Legendre and Zernike moments as they are all global moments. This means that, by using Jacobi moments, local information at a particular position of the image can be extracted. Experimental results are given to support these claims.