A Review of Evaluation of Optimal Binarization Technique for Character Segmentation in Historical Manuscripts

Abstract

Abstract-A number of binarization techniques have been proposed in the past for automatic document processing. Although some studies have aimed to evaluate the performance of binarization algorithms, there is no automatic system that is capable of selecting the most appropriate method of binarization. While preprocessing techniques can be applied, binarization is essential to extract the objects in the first place before the characters can be separated for recognition. Although there are several commonly used binarization approaches, there is no single algorithm that is suitable for all images. Hence, there is a need to determine the optimal binarization algorithm for each image. The objective of this paper is to present a survey of the existing methods of binarization and evaluation measurement which have been developed recently. This will lead to the proposal and development of an approach for automatic selection of binarization techniques in handling historical document images.

Study on feature extraction methods for character recognition of Balinese script on palm leaf manuscript images

“…The network considers the value of each pixel of the input image as the input layer. In this experiment, we used Tensorflow library 2 .…”

“…For example, the IHCR challenges for historical documents discovered in the palm leaf manuscripts, written using the script originating from the region of Southeast Asia, such as Indonesia, Cambodia, and Thailand. Ancient palm leaf manuscripts from Southeast Asia have received great attention from historian and also researchers in the field of document image analysis, for example some works on document analysis of palm leaf manuscript from Thailand [1,2] and from Indonesia [3,4]. The physical condition of the palm leaf documents is usually degraded.…”

A New Scheme for Text Line and Character Segmentation from Gray Scale Images of Palm Leaf Manuscript

“…Unfortunately, for some type of historical document image, for example the palm leaf manuscript images from Southeast Asia, the binarization process to separate the ancient text from the background is a real challenge [14][15][16]. A review of evaluation of optimal binarization technique for character segmentation in historical manuscripts was presented in [14]. In our previous work [15], we already experimented and compared several alternative well-known binarization algorithms on the palm leaf manuscript images.…”