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.

Visual Gesture Character String Recognition by Classification-Based Segmentation with Stroke Deletion

Abstract: The recognition of character strings in visual gestures has many potential applications, yet the segmentation of characters is a great challenge since the pen lift information is not available In this paper, we propose a visual gesture character string recognition method using the classification-based segmentation strategy In addition to the character classifier and character geometry models used for evaluating candidate segmentation-recognition paths, we introduce deletion geometry models for deleting stroke segments that are likely to be ligatures To perform experiments, we built a Kinect-based fingertip trajectory capturing system to collect gesture string data Experiments of digit string recognition show that the deletion geometry models improve the string recognition accuracy significantly The string-level correct rate is over 80%

Generalization of Hindi OCR Using Adaptive Segmentation and Font Files

Abstract: In this chapter, we describe an adaptive Indic OCR system implemented as part of a rapidly retargetable language tool effort and extend work found in [20, 2]. The system includes script identification, character segmentation, training sample creation, and character recognition. For script identification, Hindi words are identified in bilingual or multilingual document images using features of the Devanagari script and support vector machine (SVM). Identified words are then segmented into individual characters, using a font-model-based intelligent character segmentation and recognition system. Using characteristics of structurally similar TrueType fonts, our system automatically builds a model to be used for the segmentation and recognition of the new script, independent of glyph composition. The key is a reliance on known font attributes. In our recognition system three feature extraction methods are used to demonstrate the importance of appropriate features for classification. The methods are tested on both Latin and non-Latin scripts. Results show that the character-level recognition accuracy exceeds 92% for non-Latin and 96% for Latin text on degraded documents. This work is a step toward the recognition of scripts of low-density languages which typically do not warrant the development of commercial OCR, yet often have complete TrueType font descriptions.

Character recognition system

Abstract: An optical character recognition system which automatically reads handwritten characters and the like which do not have to be printed in a special format. Recognition tables derived from the pattern bit map and from a skeleton pixel matrix describe the character in terms of the relative position of the pixels along the borders of the character and provide a plurality of recognition strings, one in each table, for the front and rear views of the character and for the shape of the holes in the character which are opened from the top (as in the numeral four) or opened from the bottom (as in the numeral seven). From the recognition tables, the characters are recognized by searching recognition files containing blocks of successions of lines of code corresponding selectively to the codes in the recognition tables. The recognition file is arranged in hierarchal order so that the blocks in the file which represent characters having the lowest level of recognition difficulty in the character set to be recognized are searched first, the next highest level next and so forth. Recognition blocks for the next character in the group of blocks for the same difficulty of recognition level or to blocks for the next level of recognition difficulty. In this manner characters are recognized with a high degree of reliability and an indication of failure to recognize the character occurs rather than misrecognition.

Handwritten Arabic word recognition based on Ridgelet transform and support vector machines

Abstract: We propose a method for handwritten Arabic word recognition based on the combination of the Ridgelet transform and SVMs. Ridgelets are used for generating pertinent features of handwritten words while the classification stage is based on the One-Against-All multiclass implementation of SVMs. The experimental investigation is conducted on a vocabulary of twenty-four words extracted from the IFN/ENIT database. The Ridgelet performance is assessed comparatively to the results obtained for Radon and uniform grid (zoning) features. Results highlight the reliability of the Ridgelet-SVM combination for handwritten Arabic word recognition.

Segmentation and recognition of handwritten dates

Abstract: Presents an HMM-MLP hybrid system to recognize complex date images written on Brazilian bank cheques. The system first segments implicitly a date image into sub-fields through the recognition process based on an HMM-based approach. Afterwards, the three obligatory date sub-fields are processed by the system (day, month and year). A neural approach has been adopted to work with strings of digits and a Markovian strategy to recognize and verify words. We also introduce the concept of meta-classes of digits, which is used to reduce the lexicon size of the day and year and improve the precision of their segmentation and recognition. Experiments show interesting results on date recognition.