Image Analysis Using Optimized Generalized Legendre Moments Invariants and Artificial Bee Colony (ABC) Algorithm

In this work, we proposed a new algorithm for optimizing the parameter generalized Legendre moments invariants using the artificial bee colony (ABC) optimization algorithm and support vector machines (SVM) to analysis the performance of reconstruction and classification of images, where we can derive the new set of invariants based on the geometric basis and the generalized Legendre polynomials. However, it was very important to optimize the parameter in order to obtain a good result. We presented a systematic selection method to find the optimal values ​​of fractional parameters for image analysis applications. Equally important, we introduced an adaptive scheme defining fractional parameters based on the characteristics of the image. The use of the (ABC) algorithm to determine the optimal parameter of Legendre polynomials (LPs), this which improves the quality of the of the reconstruction. The results obtained show, on the one hand that the method based on the ABC algorithm offers better image reconstruction on the other hand, the comparison with other algorithms indicates the power of the proposed method. Finally, the method proposed is inexpensive in computationally and very fast, it is useful in several computer vision applications.