Abstract— In all the modern security systems biometric plays an important role. Among the existing biometric methods, fingerprint recognition is easiest and low cost method. The storing of the captured fingerprint images is the main problem. Wavelet Scalar Quantization (WSQ) is the first technique proposed to compress fingerprint images and it provides only low values of compression ratio and also low quality of the compressed image. This paper proposed a unique method called sparse representation to provide better compression ratio and better quality of the compressed image than WSQ technique. In the proposed method, the dictionary is constructed for the predefined fingerprint image patches and are determined by l1-minimization. These dictionary patches are quantize, encoded and are used for the calculation of patches of new fingerprint images. The pre and post compression factor provides the difference between patches. The performance of the proposed method is evaluated by the computation of PSNR and AFIS. The proposed method is gives better compression ratio and better quality than the existing WSQ method. This method is mainly used for the high quality compression ratio of the fingerprint images.

Index Terms— Biometrics, Compression techniques, AFIS, WSQ

Cite: R Damodara Varaprasad and Chandra Mohan Reddy Sivappagari, "IMPLEMENTATION OF HIGH EFFICIENT COMPRESSION METHOD FOR FINGERPRINT IMAGES," International Journal of Electrical and Electronic Engineering & Telecommunications, Vol. 4, No. 4, pp. 7-14, October 2015.