Low Pass Filter-Based Enhancement of Arabic Handwritten Document Images

Document image processing gained due attention during the recent few decades as a result of technological advancements in the twenty-first century pertaining to facilitating the enhancements of the pictorial data for human interpretation. During the image acquisition process, some unwanted information needs to be removed by using several preprocessing techniques. Among those, filtering helps foster the image processing through enhancing and noise removal. This paper aims to demonstrate the low pass filter which is also used in Fourier and wavelet transformations. Through applying the low pass on a specifically built database, it was found out that low pass preserves the edge details and denoises the image by preserving details, and also it produces smooth and blurs document images. In this work, we conducted experimentation on 100 images of our own handwritten dataset, and the proposed method is evaluated based on the quantitative measures.