An Analysis of Image Compression Using Neural Network

Image compression belongs to the area of data compression because the image is itself made up of data and the task of compressing images has become vital in our current life. Because the scenario is that images are required to build more attractive contents, also in today’s world smartphones cover large fraction of internet traffic and are having low data bandwidth on average. Due to these factors and restrictions on bandwidth and other computing capabilities it has necessary for developers of websites/applications to reduce either size or resolution or both of an image to improve responsiveness of your websites/apps. For this purpose image compression is divided into two categories these are lossless image compression and lossy image compression. The requirement for lossless image compression is that during the decompression process the image data must be recovered without/with negligible loss in image quality while in lossy image compression certain amount or level of error is allowed in image data to achieve better compression ratios and performance. Neural Networks because of their good performance have been used to implement the task of image compression and there are multiple modified neural networks that are proposed to perform image compression tasks, however the consequent models are big in size, require high computational power and also best suited for fixed size compression rate and some of them are covered in this survey report.