Hybrid Fingerprint Image Compression Technique Using Wavelet Transformation and Block Truncation Coding with Rle

The Hybrid technique that combines the Wavelet Transformation and Block Truncation Coding is a new framework proposed for fingerprint image compression. The proposed technique includes image enhancement system, forward DWT, encoding system, decoding system and inverse DWT. Image enhancement techniques like histogram equalization and wiener filtering has been performed on the image before Wavelet transformation so that the noisy image can also be retrieved fast. Wavelet based compression are largely used due to the competitive compression ratios that can be achieved at high quality without much blocking artefacts. The BTC converts a gray level image into a binary image with a predefined threshold. The coded bits are written into separate text file as sequence of symbol pairs (run, value) using runlength coding. Experiments on an image database of grayscale JPEG (uncompressed) images shows that the proposed technique performs well in compression and decompression. The Quality metrics used are CR, SNR, MSE and PSNR. General terms AFIS, FBI, MSE, PSNR, CR, SNR