Abstrakt

Ancient Degraded Document Binarization Using Mean Shift Technique

P.S. Jonesherine, J. Paul Richardson Gnanaraj

In this paper a Mean shift algorithm is employed for ancient document images is proposed, as well as a post processing method that can improve any Binarization method. We introduce a local-global Mean Shift based colour image segmentation approach. It is a two-steps procedure carried out by updating and propagating cluster parameters using the mode seeking property of the global Mean Shift procedure. The first step consists in shifting each pixel in the image according to its R-Nearest Neighbour Colours (R-NCC) in the spatial domain. The second step process shifts only the previously extracted local modes according to the entire pixels of the image. Binarized model is made efficient by including mean shifting technique in the image. While in the post processing step, specialized adaptive Gaussian and median filters are considered. The result shows the output binarized image with the removal of global bleed through and few other degradation and proposed method is more efficient and provide better computed PSNR values comparing to the prior art.