Handwritten Character Recognition Using Tchebychev Moments and Pseudo-Zernike Moments

スポンサーリンク

概要

• 論文の詳細を見る

• This paper deals with handwritten character recognition using Tchebychev moments and pseudo-Zernike moments as the input data into classifiers by means of the neural network. These two types of moments are the orthogonal moments. The orthogonal moments are divided roughly into two types, i.e. the orthogonal real moments and the orthogonal complex moments, and there exist several kinds of moments in each individual group. Tchebychev moments are ones newly proposed as the orthogonal real moments using Tchebychev polynomials. In the meanwhile, pseudo-Zernike moments are the orthogonal complex moments which are conducted by using new method in recognition experiments. In this paper, it is shown that Tchebychev moments and pseudo-Zernike moments are respectively excellent in each group. The comparison between Tchebychev moments and pseudo-Zernike moments was attempted. As the result, it is explained that pseudo-Zernike moments is sligtly superior to Tchebychev moments in the recognition ability.

• 東海大学の論文

著者

• Kawabe Hidekazu

  Department Of Electronics

関連論文

• Rotation-invariant Digital Pattern Recognition Using Singular Value Decomposition Applied Complex-log Mapping
• Handwritten Character Recognition Using Tchebychev Moments and Pseudo-Zernike Moments
• Recognition of Rotated Images Using Complex-log Mapping
• Rotation-Invariant Character Recognition by the Moments Using Associated Legendre Functions
• Recognition of Rotated Images Using Singular Value Decomposition Applied to Complex-log Mapping
• Character Recognition by a Neural Network Using Pseudo-Zernike Moments as Input Data

スポンサーリンク