Handwritten Image Enhancement Based on Neutroscopic‐Fuzzy and K‐Mean Clustering
Jaspreet Kaur, Divya Gupta, Simarjeet Kaur, Amrinder Pal Singh · 2024
Image processing is the process of applying different operations to an image in order to either improve it or extract useful information from it. In this chapter, we are addressing the problems that are associated with handwritten document with the passage of time. Hand-written documents can deteriorate for a number of reasons, including the original document's quality, the ink or paper used, and the document's age. As a result, the text could become unclear, blurry, or faded, making it challenging to read and comprehend. Identification and recognition of old manuscripts is very difficult from low resolution images due to poor boundary and contrast. Work done in this field deals with recognizing content of old manuscripts. Many documents that are produced today, such as personal letters, pictures, agreements, newspapers and medical records serve as valuable historical documents for some people in the future. Further, most of them will be vanished in the future because many documents are often made from faulty materials that disappear, tear or destroy over time or are even stored in digital formats then lost track over time. Preservation of ancient manuscripts against degradation is one of the main works of the library. In this slog, we try to enhance the handwritten document that has been degraded with passage of time using various segmentation and clustering techniques. For artifact fortification collective approach neutroscope and fuzzy type is used to increase the readability of the document. This chapter offers a thorough analysis of the techniques for restoring antique manuscripts with damaged backgrounds. Three different types of enhancement techniques were found, including (a) local threshold image enhancement, (b) k-means clustering, and (c) combination of neutroscopic set (NS) and fuzzy type-1 image enhancement. Images used for enhancement can be any scanned image or camera-captured images. This work employs image segmentation methods for separating foreground and background and clustering techniques for enhancing the quality of image.