Digit Classification on Signboards for Telephone Number Recognition

Authors:
Takuma Yamaguchi;Yasuaki Nakano;Minoru Maruyama;Hidetoshi Miyao;Toshihiro Hananoi
Affiliations:
-;-;-;-;-
Venue:
ICDAR '03 Proceedings of the Seventh International Conference on Document Analysis and Recognition - Volume 1
Year:
2003

Citing 5
Cited 3

Skew Angle Detection of Digitized Indian Script Documents

IEEE Transactions on Pattern Analysis and Machine Intelligence
Skew and Slant Correction for Document Images Using Gradient Direction

ICDAR '97 Proceedings of the 4th International Conference on Document Analysis and Recognition
Color Texture-Based Object Detection: An Application to License Plate Localization

SVM '02 Proceedings of the First International Workshop on Pattern Recognition with Support Vector Machines
Information Capturing Camera and Developmental Issues

ICDAR '99 Proceedings of the Fifth International Conference on Document Analysis and Recognition
Extraction of Place-Name from Natural Scenes

IWFHR '02 Proceedings of the Eighth International Workshop on Frontiers in Handwriting Recognition (IWFHR'02)

Improved Text-Detection Methods for a Camera-based Text Reading System for Blind Persons

ICDAR '05 Proceedings of the Eighth International Conference on Document Analysis and Recognition
A genetic framework using contextual knowledge for segmentation and recognition of handwritten numeral strings

Pattern Recognition
Global Localization for the Mobile Robot Based on Natural Number Recognition in Corridor Environment

Neural Information Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a digits classification system to recognizetelephone numbers written on signboards. Candidateregions of digits are extracted from an image throughedge extraction, enhancement and labeling. Since the digitsin the images often have skew and slant, the digits are recognizedafter the skew and slant correction. To correct theskew, Hough transform is used, and the slant is correctedusing the method of circumscribing digits with tilted rectangles.In experiments, we tested a total of 1,332 images ofsignboards with 11,939 digits. We obtained a digit extractionrate of 99.2% and a correct digit recognition rate of98.8%.