Illumination Reduction for Camera Captured Documents Using Block Processing and Homomorphic Filtering Technique

Abstract

When acquiring document image using portable camera like mobile and other devices like HIDs in uncontrolled environment often, exhibits various forms of distortions to captured image, Which in turn degrades the quality of the image. Data quality of the image is highly affected by illumination. Therefore removal of such illumination and increase of data quality is necessary for human perception and machine recognition. In this paper, We propose a novel methodology for Illumination correction and data quality improvement for camera captured documents. The proposed methodology uses Homomorphic and Morphological filtering techniques to reduce illumination and some efficient binarization and enhancement techniques to improve data quality. Experimental results on camera captured documents shows that results obtained are encouragingly reliable, effective and efficient compared with the existing proven methods.