Chapter 4 - Statistical measures for Palmprint image enhancement

Abstract

Image enhancement aims to process an image so that the resulting image is of more enhanced quality as compared to original image. With special regard to various filtering and statistical techniques, this article focuses on statistical measures for palmprint image enhancement. For the estimation of the quality of improved images, statistical analysis of image quality measurements can be performed. Palmprint image enhancement as a preprocessing technique becomes very relevant as well as challenging for biometric recognition systems. The ultimate target of palmprint image enhancement is to produce an image of better quality. In the quality of the high spatial resolution palmprint recognition system, palmprint image consistency is an important factor. There is also a limited amount of work specifically related to image enhancement techniques for palmprint images. Various types of statistical analysis for palmprint image enhancement techniques are discussed in this chapter. Mainly, the discussion is based on image quality measurement, both subjective and objective. In this chapter, focus was made on the objective quality measurement which is based on the statistical analysis. Such as normalized cross correlation (NK), mean square error (MSE), average difference (AD), peak signal-to-noise ratio (PSNR), maximum difference (MD), and structural content (SC). Such study will be helpful in understanding the most suitable statistical measures for palmprint image enhancement assessment. The comparative analysis is presented in the work.