Cleaning of Ancient Document Images Using Modified Iterative Global Threshold

Ancient document Image processing is an important area attracting many researchers in the recent period. Binarization is the first step while cleaning the document for further processing. Based on the degradation of the original document, either global or local thresholding methods are preferred. Thresholding phenomenon is a simple and practical approach to identify the cluster of pixels that are most likely associated with background information, while separating the object information. In this paper we propose a modified iterative global thresholding approach to separate the clusters of fore ground and background. The relative closeness towards background intensity is computed in each iteration after image equalization. Camera captured images of ancient printed documents, stone carvings and palm leaves are evaluated in the present paper.