Intelligent word recognition

About: Intelligent word recognition is a research topic. Over the lifetime, 2480 publications have been published within this topic receiving 45813 citations.

Removal of Obstacles in Devanagari Script for Efficient Optical Character Recognition

Abstract: There are different defining problems are still need to resolve for the high accuracy of Optical Character Recognition (OCR) of Devanagari script. The more prominent number of font availability in Devanagari script is a challenge for the character detection in India. Therefore to conquer this problem, here is proposed a system for identification of diverse font of characters which helps to improve the OCR system accuracy. The adequacy technique for recognition of character for number of font styles in Hindi script is shown by this methodology. The proposed methodology is suitable for handling script styles information for Devanagari script is indicated by the outcomes. A procedure for enhancing the recognition accuracy of Devanagari OCR System by creating idea for discovery emphasis words such as Bold, Italic and underline words is exhibited in this work. We have taken most generally utilized Devanagari font styles, for example, kruti Dev 714, DevLys 240, and Alekh for our benchmark testing. Recognition of text style in Hindi script record enhances the execution of Hindi OCR framework.

Segregated Handwritten Character Recognition using GLCM Features

Abstract: Handwritten document recognition is an area of pattern recognition that has been showing impressive performance in the machine printed text. Handwritten document recognition is an intricate task to various writing styles of individual person. The system first identifies the contour in a handwritten document for segmentation and features are extracted from the segmented character. This paper uses GLCM(Gray Level Co-occurrence Matrix) for character recognition. Features of a character has been computed based on calculating the pairs of pixel with specific values and specified spatial relationship occurrence in an image. First order and second order textures are used to measure the intensity of the original pixels. Data were collected from dierent persons, and the system is trained using SVM with various writing styles. The proposed system achieves a maximum recognition accuracy of 95.2% with training and testing data using GLCM as features and SVM with RBF kernel function.

Arabic script based character segmentation: A review

Abstract: Segmentation based Arabic script based languages character recognition has been a popular field of research for many years. The challenging nature of Arabic script recognition has attracted the attention of researchers from both industry and academic circles but these efforts have not achieved good results until now. Segmentation of Urdu script when written in Nasta'liq writing style is very difficult task due to the complexity of writing style as compare to Naskh writing style. Good segmentation is one of reasons for high accuracy. Character segmentation has been a critical phase of the OCR process. The higher recognition rates for isolated characters as compare to results of words or connected character well illustrate the importance of segmentation. Current study investigate the recent work for character segmentation and challenges for segmentation for Arabic script based languages.

Selective Attention for Handwritten Digit Recognition

Abstract: Completely parallel object recognition is NP-complete Achieving a recognizer with feasible complexity requires a compromise between parallel and sequential processing where a system selectively focuses on parts of a given image, one after another Successive fixations are generated to sample the image and these samples are processed and abstracted to generate a temporal context in which results are integrated over time A computational model based on a partially recurrent feedforward network is proposed and made credible by testing on the real-world problem of recognition of handwritten digits with encouraging results

Benefit of multiclassifier systems for Arabic handwritten words recognition

Abstract: In order to improve the results of single classifiers, the study of multiple classifier systems has become an area of intensive research in pattern recognition. In this paper, two types of features are fed to a number of artificial neural networks (ANN). Then, their respective responses are combined for the recognition of handwritten Arabic literal words. Different parallel combination schemes are presented, including the use of an ANN as a meta classifier. Their results are then compared and conclusions on the most suitable approach are drawn.