AN EFFICIENT BINARIZATION ON DEGRADED MALAYALAM HANDWRITTEN PALM LEAF MANUSCRIPT DATASET ENHANCEMENT USING 3D - DS CLAHE ALGORITHM

Abstract

Research Area on Degraded Palm Leaf Manuscripts is a tedious task for document processing. The Proposed work has taken a dataset on the Degraded Malayalam Palm Leaf Manuscript (DMPLM). Degrading is similar to damage and deterioration that results in uneven illumination, decolorization, holes, cracks, moisture, staining, physical damage, insect bite, rodent activity, splitting, and fractionalization. The Process involves solving the complexity behind the density adjustments on the captured dataset. The Proposed work resolves complexity by applying 3D depth slicing in the CLAHE (Contrast Limited Adaptive Histogram Equalization) algorithm with a Cumulative Distribution Function (CDF) component. The proposed work shows a better result score of 98% in consecutive Efficient Recognition of Text Binarization process compared to existing methods