Title: An optimising method of Japanese character recognition based on improved support vector machine

Authors: Yang Wang; Pan Zhao

Addresses: College of Oriental Language, Harbin Normal University, Harbin, 150000, China ' College of Oriental Language, Harbin Normal University, Harbin, 150000, China

Abstract: Aiming at the problem of low recognition accuracy and long recognition time due to poor feature extraction, a Japanese character recognition method based on improved support vector machine was studied. First, the mean filtering method is used to denoise Japanese text samples, and then Gabor filter is used to extract features, and Fourier transform is used to optimise the extraction speed. Then, the initial text recognition method is constructed by introducing support vector machine and improving the penalty coefficient algorithm. Finally, Japanese character recognition is realised by using the multi-layer perceptron function as the kernel by processing linear indivisible samples with the relaxation term minimisation. The experimental results show that the proposed method can effectively extract Japanese character features with a recognition accuracy of 97.4% and a recognition time of only 1.3 s, which effectively solves the problems of low recognition accuracy and long recognition time.

Keywords: support vector machine; Japanese characters; identification method; Gabor filter; Fourier transform; penalty coefficient; Kernel function.

DOI: 10.1504/IJRIS.2025.148712

International Journal of Reasoning-based Intelligent Systems, 2025 Vol.17 No.5, pp.291 - 300

Received: 30 May 2023
Accepted: 14 Jul 2023
Published online: 21 Sep 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article