Abstract
AbstractAn important problem in image recognition is extracting rotation in invariant features. Currently, a powerful technique based on Zernike moments is applied when the moments are used to address this problem. In this paper, we use associated Legendre functions of the first kind to derive moments and attempt to extract rotation invariant features based on them. These new moments are complex data as are Zernike moments and their basis functions are orthogonal functions. The images reconstructed from these moments demonstrated high quality than when using Zernike moments. Furthermore, evaluation functions suited to these moments are designed and image recognition tests are performed by several computer simulations. The results demonstrate the strong recognition ability for images degraded by noise or distortion, as well as illustrate features like recognition in regions of low degrees.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
