A STUDY ON IMAGE COMPRESSION WITH NEURAL NETWORKS USING MODIFIED LEVENBERG METHOD

Authors

  • Dr. K.PREMA

  • N.SREEKUMAR

Keywords:

Image complexity, PSNR, Levenberg-Marquardt, Multi-layer neural network

Abstract

In this paper, an adaptive method for image compression that is subjective on neural networks based on complexity level of the image. The multilayer perceptron artificial neural network uses the different Back-Propagation artificial neural networks in processing of the image. The original images taken, for instance 256*256 pixels of bitmap image, each block of image into one network selection, according to each block the value of pixels in image complexity value is calculated. To estimate each value of the images in a block can be evaluated and trained. Best PSNR in selecting images to be compressed with a modification Levenberg-Marquart for MLP neural network is taken. The algorithm taken a good research of result to each block of image. The taken time reduces the learning procedure for running each block of images. Finally, a neural network taken for the Back Propagation artificial neural network.

How to Cite

Dr. K.PREMA, & N.SREEKUMAR. (2011). A STUDY ON IMAGE COMPRESSION WITH NEURAL NETWORKS USING MODIFIED LEVENBERG METHOD. Global Journal of Computer Science and Technology, 11(3), 1–6. Retrieved from https://computerresearch.org/index.php/computer/article/view/673

A STUDY ON IMAGE COMPRESSION WITH NEURAL NETWORKS USING MODIFIED LEVENBERG METHOD

Published

2011-01-15