Volume : III, Issue : X, October - 2014

Robust Image Normalization Using H–Connection Technique for Face Recognition

K. M. Ponnmoli, Dr. S. Selvamuthukumaran

Abstract :

Normalization process for face recognition process is a vital one that can be done by perform the H–connection of pixel coordinates from badly degraded face images are very confronting tasks because of high variation between the image background and the foreground layers of different document images. In this proposed system the novel image normalization using the H–CON (Hierarchical Connection) technique handle these issues with the adaptive image contrast. Combination of the local image contrast and the local image gradient constitutes the adaptive image contrast. It is adaptive to the variation of text and background with the image degradation. In the proposed system the map is first developed for the degradation of the image. The contrast map is then binarize and it is integrated with the Canny’s edge map to recognize the pixel edges. The feature of the image is divided by a threshold and it is determined with the intensities in the pixel edges. In the proposed system the method is clear and strong and includes minimum boundary tuning. The experiments on the test bed has a several challenging tasks and documents .Estimated with other techniques shows the PSNR,FAR and noise level of the output image.

Keywords :

Article: Download PDF   DOI : 10.36106/ijsr  

Cite This Article:

K.M.Ponnmoli, Dr. S. Selvamuthukumaran Robust Image Normalization Using H-Connection Technique for Face Recognition International Journal of Scientific Research, Vol : 3, Issue : 10 October 2014

Number of Downloads : 1195

References :