Applying Domain Knowledge to the Recognition of Handwritten Zip Codes

We present a simple system that exploits domain knowledge to improve the segmentation and recognition of handwritten ZIP codes. Specifically, we show that the concept of metaclasses of digits, introduced by Morita et al. [16] for recognition of Brazilian bank check dates, can be extended to ZIP code recognition. We also show that, when this domain knowledge is present, integrated segmentation and recognition is not required for the recognition of handwritten ZIP codes, as claimed by Liu et al. [4].

[1]  Michael R. Scheessele,et al.  Human Performance on the USPS Database , 2007 .

[2]  Ching Y. Suen,et al.  Recognition of courtesy amounts on bank checks based on a segmentation approach , 2002, Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition.

[3]  Patrice Y. Simard,et al.  Best practices for convolutional neural networks applied to visual document analysis , 2003, Seventh International Conference on Document Analysis and Recognition, 2003. Proceedings..

[4]  Ching Y. Suen,et al.  A trainable feature extractor for handwritten digit recognition , 2007, Pattern Recognit..

[5]  Flávio Bortolozzi,et al.  Segmentation and recognition of handwritten dates: an HMM-MLP hybrid approach , 2003, Document Analysis and Recognition.

[6]  Cheng-Lin Liu,et al.  Handwritten digit recognition: benchmarking of state-of-the-art techniques , 2003, Pattern Recognit..

[7]  Hiroshi Sako,et al.  Integrated segmentation and recognition of handwritten numerals: comparison of classification algorithms , 2002, Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition.

[8]  Teuvo Kohonen,et al.  The self-organizing map , 1990, Neurocomputing.

[9]  Peter Norvig,et al.  Artificial Intelligence: A Modern Approach , 1995 .

[11]  Hiroshi Sako,et al.  Effects of classifier structures and training regimes on integrated segmentation and recognition of handwritten numeral strings , 2004, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[12]  Hiroshi Sako,et al.  Performance evaluation of pattern classifiers for handwritten character recognition , 2002, International Journal on Document Analysis and Recognition.

[13]  Djamel Bouchaffra,et al.  Recognition of strings using nonstationary Markovian models: an application in ZIP code recognition , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[14]  E. Sackinger,et al.  Neural-Network and k-Nearest-neighbor Classifiers , 1991 .

[15]  Ching Y. Suen,et al.  Automatic segmentation and recognition system for handwritten dates on Canadian bank cheques , 2003, Seventh International Conference on Document Analysis and Recognition, 2003. Proceedings..

[16]  Azriel Rosenfeld,et al.  Computer Vision , 1988, Adv. Comput..

[17]  Yann LeCun,et al.  The mnist database of handwritten digits , 2005 .

[18]  K. Mohiuddin.  International Conference On Document Analysis and Recognition , 1995, Proceedings of 3rd International Conference on Document Analysis and Recognition.

[19]  Hiroshi Sako,et al.  Handwritten digit recognition using state-of-the-art techniques , 2002, Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition.

[20]  Linda G. Shapiro,et al.  Computer Vision , 2001 .