Threshold Approach to Handwriting Extraction in Degraded Historical Document Images

Handwriting extraction is the skill of a system to get and translate comprehensible hand written input via sources such as document, photos, tough screen and other devices. The picture of the written document is used to detect written text by the use of optical scanning i.e. known as optical character recognition. Handwriting extraction basically uses optical character recognition. Conversely, an absolute hand writing extraction process that handles format and perform correct segmentation into typescript and searches for the most reasonable terms. Handwriting extraction is a process of automatic typesetting of text from a picture to letter sets that are exploitable by a system or a computer by the use of textprocessing software. The information received via this method form is treated as static illustration of hand writing. Off line handwriting recognition is relatively complex due to the reason that different persons have differences in the handwriting styles. Today, Optical Character Recognition engines mainly focus on instrument printed text and Intelligent Character Recognition for hand written text. The proposed system uses the above mentioned key features with going one step further. One of the most impressive aspects of human visual processing is the ability to recognize objects despite severe degradations in image quality. The paper focuses on the recognition of impoverished handwritten documents. General Terms Optical Character Recognition (OCR), Intelligent Character Recognition (ICR), Binarization, Ocular Scanning (optical character extraction)