Extraction of Nth‐order features of line patterns considering redundant information
Norifusa Nishiyama, Kazuo Inoue · Systems and Computers in Japan · 1987
Abstract The extraction of geometric features is a basic and significant process for pattern recognition. This paper proposes a method of extracting geometric feature by referring redundant information around the feature point. For written or oral information, even if some information is unable to be understood, we can infer meaning by knowing redundant information. Using the same principle as mentioned in the opening paragraph, this paper introduces a method to extract the geometric feature by detecting redundant information around the feature point. Finally, to verify the usefulness and effectiveness of this method, computer simulation experiment is performed on test patterns and handwritten numerals.