Image compression is considered as application performed for compression of data in digital format images. Digital images are comprised with large amount of information that requires bigger bandwidth. The techniques of image compression can be generally categorized into two types: lossless & lossy technique. DCT (discrete cosine transform) can also be used for compressing an image and also approaches like Huffman encoding, quantization & such steps are required for compression of images with JPEG format. The format of JPEG can be used for both of the RGB (colored) & YUV (gray scale) images. But here our main concentration is over decompression & compression of gray scale images. Here 2D-DCT can be used for transformation of a 8×8 matrix of images to an elementary frequency elements. The DCT is considered to be a mathematical function which will transform an image of digital format from spatial to frequency domain. It is very much easy to implement Huffman encoding & decoding for minimizing the complexity of memory. In this proposed technique, the analog image pixels are transformed to discrete image pixel, and therefore compression is preformed. On the receiving side, the pixels are decompressed for obtaining the actual image. The PSNR is computed for analyzing the quality of image.

Keywords : JPEG, DCT, IDCT, Huffman Image Technique, PSNR.