Modeling of neural image compression using gradient decent technology

--------------------------------------------------------------ABSTRACT------------------------------------------------------In this paper, we introduce a new approach for image compression. At higher compression rate, the decompressed image gets hazy in GIF and PNG, an image compression technique. In order to overcome from those problems artificial neural networks can be used. In this paper, gradient decent technology is used to provide security of stored data. BP provides higher PSNR ratio, with fast rate of learning. Experiments reveal that gradient decent technology works better than Genetic algorithm where the image gets indistinguishable as well as not transmitted in a secured manner.

[1]  J. Jiang,et al.  Image compression with neural networks - A survey , 1999, Signal Process. Image Commun..

[2]  Khashayar Khorasani,et al.  Adaptive Constructive Neural Networks Using Hermite Polynomials for Image Compression , 2005, ISNN.

[3]  Chao Zhang,et al.  Improved Variance-Based Fractal Image Compression Using Neural Networks , 2006, ISNN.

[4]  Swanirbhar Majumder,et al.  Digital Image Compression Using Neural Networks , 2009, 2009 International Conference on Advances in Computing, Control, and Telecommunication Technologies.

[5]  Reza Askari Moghadam,et al.  Image compression using linear and nonlinear principal component neural networks , 2009, 2009 Second International Conference on the Applications of Digital Information and Web Technologies.

[6]  Yevgeniy V. Bodyanskiy,et al.  Fast Training of Neural Networks for Image Compression , 2010, ICDM.

[7]  John W. Woods,et al.  Digital Image Compression , 2012 .