Adaptive thresholding of document images based on Laplacian sign

We present a new technique for document image binarization to manage different situations in an image. The problems of image binarization caused by illumination, contrast, noise and much source type-related degradation are addressed. A new technique is applied to determine a local threshold for each pixel. The idea of our technique is to update locally the threshold value whenever the Laplacian sign of the input image changes along the raster scanned line. The Differential of Gaussian (DoG) is used to define the sign image. The proposed technique is tested with images including different types of document components and degradations. The results were compared with a global thresholding technique. It is shown that the proposed technique performs well and is highly robust.