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.

A Multi-stream Approach to Off-Line Handwritten Word Recognition

Abstract: We present in this paper a new approach based on multi-stream hidden Markov models (HMM) for the recognition of off-line handwriting. Every word is presented by two HMM models: the first one is learned with features extracted from upper contour, the second with features extracted from lower contour. The combination of these two sources of information is studied using the multi-stream framework. We present experiment results obtained on a database composed of isolated words extracted from incoming mail documents.

Spiral recognition methodology and its application for recognition of Chinese bank checks

Abstract: This paper presents the spiral recognition methodology with its application in unconstrained handwritten Chinese legal amount recognition in a practical environment of a CheckReader/spl trade/. This paper first describes the failed application of neural network - hidden Markov model hybrid recognizer on Chinese bank check legal amount recognition, and explains the reasons for the failure: the neural network - hidden Markov model hybrid recognizer cannot handle the complexity in the training for Chinese legal amounts. Then a spiral recognition methodology is presented. This methodology enables the system to increase its recognition power (both the recognition rate and the number of recognized characters) during the training iterations. Some experiments were done to show that the spiral recognition methodology has a high performance in the recognition of unconstrained handwritten Chinese legal amounts. The recognition rate at the character level is 93.5%, and the recognition rate at the legal amount level is 60%. Combined with the recognition of courtesy amount, the overall error rate is less than 1%.

A syntax directed level building algorithm for large vocabulary handwritten word recognition

Abstract: This paper describes a large vocabulary handwritten word recognition system based on a syntax{directed level building algorithm (SDLBA) that incorporates contextual information. The sequences of observations extracted from the input images are matched against the entries of a tree{structure lexicon where each node is represented by a 10{state character HMM. The search proceeds breadth| rst and each node is decoded by the SDLBA. Contextual information about writing styles and case transitions is injected between the levels of the SDLBA. An implementation of the SDLBA together with a 36,100{entry lexicon is described. In terms of recognition speed, the results show that the SDLBA together with the tree{structured lexicon outperforms a baseline system that uses a Viterbi{ at{lexicon scheme while maintaining the same accuracy and consuming a reasonable amount of memory.

On-line handwritten Chinese word recognition based on lexicon

Abstract: We proposed a new method for on-line handwritten Chinese word recognition based on lexicon. A pre-segment strategy is first applied to the input strokes and merges the strokes into candidate character boxes. Segmentation path generation and filtering method is then invoked to generate segmentation paths from the boxes and filter out impossible paths by the recognition results of candidate character and lexicon-based technique. The optimal segmentation path along with the best matching word in the lexicon is finally chosen by a voting strategy using both recognition and geometrical information. Superiority of the proposed approach has been proven by testing it with 8550 words collected by Tablet PC. The accuracy of segmentation and recognition is about 99%

Handwritten Digit Recognition Using DCT and HMMs

Abstract: Handwritten digits recognition has been an interesting area due to its applications in several fields. Recognition of bank account numbers and zip codes are a few examples. Handwritten digits recognition is not a trivial task due to presence of large variation in writing style in available data. In order to cope with this problem both features and classifier need to be efficient. In this research, transformation based features, Discrete Cosine Transform (2D-DCT), have been used. Hidden Markov models (HMMs) have been applied as classifier. The proposed algorithm has been trained and tested on Mixed National Institute of Standards and Technology (MNIST) handwritten digits database. The algorithm provides promising recognition results on MNIST database of handwritten digits.