A quality aware technique for biometric recognition

Abstract

Noise is a very common factor for degradation of images and for better performance of any image based biometric system, enhancement of poor quality images due to noise is necessary. In most of the previously proposed techniques in the literature, it is assumed that all the images participating in the recognition process are of poor quality and an enhancement technique is blindly applied to all the images to enhance them. For example, apply Gaussian smoothing on all the images to make them noise free, however, such blind application of image enhancement degrades the quality of good images. This paper presents a quality aware technique for biometric recognition. It has utilized our previously proposed technique for image quality assessment based on noise detection. The technique first estimates the quality of the images used in the recognition and if the quality value is found to be below a certain threshold, an enhancement is applied on the image before using it for recognition. Experimental analysis has been conducted on Yale Extended Cropped Face Database. Obtained results show the effectiveness of the proposed technique.

Deep Learning Classification of Face Images with varying Illumination Conditions

“…In [14], principal components analysis (PCA), two dimensional linear discriminant analysis (2DLDA) and image similarity methods were used to classify the light source direction. Likewise, a quality-aware technique that selectively denoised only noisy images was presented by Joshi and Prakash [15]. The authors made use of image quality estimation techniques based on noise detection in order to avoid blind enhancement of images that does not require any enhancement or limited enhancement.…”

Image Quality Assessment-Based Approach to Estimate the Age of Pencil Sketch

“…Finding the age from a pencil skilled drawn from a skilled pencil sketch artist is a very challenging task [1][2][3][4][5][6][7][8][9]. Due to aging process, faces undergo continuous variations, and hence continuously updating the large databases will be a difficult task [10][11][12][13][14].…”