HMM-based Indic handwritten word recognition using zone segmentation

Abstract

This paper presents a novel approach towards Indic handwritten word recognition using zone-wise information. Because of complex nature due to compound characters, modifiers, overlapping and touching, etc., character segmentation and recognition is a tedious job in Indic scripts (e.g. Devanagari, Bangla, Gurumukhi, and other similar scripts). To avoid character segmentation in such scripts, HMMbased sequence modeling has been used earlier in holistic way. This paper proposes an efficient word recognition framework by segmenting the handwritten word images horizontally into three zones (upper, middle and lower) and recognize the corresponding zones. The main aim of this zone segmentation approach is to reduce the number of distinct component classes compared to the total number of classes in Indic scripts. As a result, use of this zone segmentation approach enhances the recognition performance of the system. The components in middle zone where characters are mostly touching are recognized using HMM. After the recognition of middle zone, HMM based Viterbi forced alignment is applied to mark the left and right boundaries of the characters. Next, the residue components, if any, in upper and lower zones in their respective boundary are combined to achieve the final word level recognition. Water reservoir feature has been integrated in this framework to improve the zone segmentation and character alignment defects while segmentation. A novel sliding window-based feature, called Pyramid Histogram of Oriented Gradient (PHOG) is proposed for middle zone recognition. PHOG features has been compared with other existing features and found robust in Indic script recognition. An exhaustive experiment is performed on two Indic scripts namely, Bangla and Devanagari for the performance evaluation. From the experiment, it has been noted that proposed zone-wise recognition improves accuracy with respect to the traditional way of Indic word recognition.Cite this as 1 @article{roy2016hmm,

Cross-language framework for word recognition and spotting of Indic scripts

“…Zone segmentation plays an important role in Indic script word recognition as proposed in [20]. As mentioned earlier, characters of most of the Indic scripts are written in upper, middle and lower zones.…”

“…With morphological combination of characters with modifiers, the number of character classes becomes huge [20]. To overcome the problem of large number of characters classes, zone-wise character segmentation and combination approach significantly reduces the combination of characters with the modifiers.…”

“…Although a number of investigations have been made towards the recognition of isolated handwritten characters and digits of Indian scripts [8,39], only a few pieces of work [7,8,17,20] exist towards offline handwritten word recognition in Indian scripts. Recognition of Indian scripts [22,38] is difficult due to their complex syntax and spatial variation of the characters when combined with other characters to form a word.…”

“…Our proposed cross-language framework uses the zone segmentation concept [20] in order to reduce the number of unique characters in Indic scripts. The major contributions of this paper are the following: 1) Use of cross language framework for word recognition and spotting: Although there are quite a few pieces of works in cross language speech recognition, the idea of cross language handwritten word recognition or spotting is a novel attempt in the handwriting recognition community.…”

“…The advantaged of zone-wise word division for cross-language similarity is explained. We have reviewed the zone segmentation method [20] in Section 3. In Section 4, we detail our proposed framework on word recognition and spotting using cross-language scripts learning.…”