The holistic paradigm in handwritten word recognition and its application to large and dynamic lexicon scenarios
IEEE Transactions on Pattern Analysis and Machine Intelligence
Algorithms for Graphics and Imag
Algorithms for Graphics and Imag
Segmentation-based recognition of handwritten touching pairs of digits using structural features
Pattern Recognition Letters
Segmentation of numeric strings
ICDAR '95 Proceedings of the Third International Conference on Document Analysis and Recognition (Volume 2) - Volume 2
Strategies in character segmentation: a survey
ICDAR '95 Proceedings of the Third International Conference on Document Analysis and Recognition (Volume 2) - Volume 2
A new system for reading handwritten zip codes
ICDAR '95 Proceedings of the Third International Conference on Document Analysis and Recognition (Volume 1) - Volume 1
Integrated segmentation and recognition of handwritten numeralswith cascade neural network
IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews
Artificial Neural Networks for Document Analysis and Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
Independent Component Analysis Segmentation Algorithm
ICDAR '05 Proceedings of the Eighth International Conference on Document Analysis and Recognition
Automatic touching detection and recognition of music chord using auto-encoding and softmax
Proceedings of the 8th International Conference on Ubiquitous Information Management and Communication
| Hi-index | 0.00 |
An approach for segmentation of handwritten touchingnumeral strings is presented in this paper. A neural networkhas been designed to deal with various types of touchingobserved frequently in numeral strings. A numeral stringimage is split into a number of line segments while strokeextraction is being performed and the segments are representedwith straight lines. Four types of primitive are definedbased on the lines and used for representing the numeralstring in more abstractive way and extracting clueson touching information from the string. Potential segmentationpoints are located using the neural network by activeinterpretation of the features collected from the primitives.Also, the run-length coding scheme is employed for efficientrepresentation and manipulation of images. On a test setcollected from real mail pieces, the segmentation accuracyof 89.1% was achieved, in image level, in a preliminary experiment.