A Novel Image Compression Using Block Local Binary Pattern (LBP) with LZW

Abstract: Image compression is aimed at reducing the data quantity, without degrading the image Quality beyond an acceptable threshold. The main advantage of LBP is it can give local texture pattern in an efficient manner. In this LBP values are used for the image compression. The description of image's local pattern results in an eight-bit binary description, but in order to restore the image from such a LBP description, the value of each central pixel is also needed. LZW coding which is simple and lossless technique is used for the LBP data compression. Finally result analysis is made based on performance parameters PSNR, MSE, SSIM.

Keywords: local binary pattern, local texture classification, Lempel-Ziv-Welch (LZW).