Training MLP Neural Network in Images Compression by GSA Method

Image compression is one of the important research fields in image processing. Up to now، different methods are presented for image compression. Neural network is one of these methods that has represented its good performance in many applications. The usual method in training of neural networks is error back propagation method that its drawbacks are late convergence and stopping in points of local optimum. Lately، researchers apply heuristic algorithms in training of neural networks. This paper introduces a new training method based on the Gravitational Search Algorithm. Gravitational Search Algorithm is the latest and newest version of swarm intelligence optimization approaches. In this algorithm، the candidate answers in search space are masses that interact with each other by gravitational force and change their positions. Gently، the masses with better fitness obtain more mass and effect on other masses more. In this research، an MLP neural network by GSA method is trained for images compression. In order to efficiency evaluation of the presented compressor، we have compared its performance toward PSO and error back propagation methods in compression of four standard images. The final results show salient capability of the proposed method in training of MLP neural networks.