Abstract: The digital image processing is exponentially increasing in today’s world which leads to storage and transmission problems. Many methods have been introduced in last decay that focuses on lossy and lossless compression. In digital network field the image compression does not tolerate the loss of data so lossless compression is preferred. Lossless compression reduces the size of image only to certain limit which is less than lossy compression. This paper compares the compression of bench mark images with various image compression algorithms that compress the images better than previous lossless image compression technique. The proposed algorithm (SEGC-N-D) is compared with DCT, DWT, Fractal, and JPEG. This process focus on identifying the non uniform areas which has high occurrence of gray level images and then segment that image using neural network. The segmented image is send to pre-compress process. In Pre-compress technique each pixel value is converted into binary and arranged in a data file as compressed using our proposed technique. Run length encoding is applied to compress the image further. To get the original image this process is reversed.

Keywords: DCT, DWT, FRACTAL, JPEG, SEGC-N-D Compression Techniques.