A novel stroke extraction method for Chinese characters using Gabor filters

✍ Scribed by Yih-Ming Su; Jhing-Fa Wang

• Publisher: Elsevier Science
• Year: 2003
• Tongue: English
• Weight: 753 KB
• Volume: 36
• Category: Article
• ISSN: 0031-3203
• DOI: 10.1016/s0031-3203(02)00086-9

✦ Synopsis

A robust stroke extraction method for Chinese characters is essential to o -line character recognition systems, which depend on stroke structure analysis to function. This study presents a novel stroke extraction method based on a directional ÿltering technique for extracting reliable and undistorted stroke segments. First, a set of Gabor ÿlters is used to break down an image of a character into di erent directional features. Next, a new iterative thresholding technique that minimizes the reconstruction error is proposed to recover stroke shape. Finally, a reÿnement process based on measuring the degree of stroke overlap is used to remove redundant stroke pieces. Experimental results show that the proposed method cannot only provide immunity against the junction-distortion and spurious-branch problems associated with a thinning-based process, but is also insensitive to shape deformation and noise.