Optimizing filter banks for supervised texture recognition

✍ Scribed by Manfred Bresch

• Publisher: Elsevier Science
• Year: 2002
• Tongue: English
• Weight: 464 KB
• Volume: 35
• Category: Article
• ISSN: 0031-3203
• DOI: 10.1016/s0031-3203(01)00075-9

✦ Synopsis

Two criteria for invariant supervised texture segmentation based on multi-channel approaches are introduced. The texture segmentation is carried out by feature extraction using multi-channel Gabor ÿltering and classiÿcation with symmetric phase-only matched ÿltering. For the feature extraction highly e cient ÿlter banks are required that enable clear distinction between feature vectors representing di erent textures in order to achieve a high classiÿcation rate. For the design of the ÿlter banks, the variances of the frequency components must be maximized. The spar hyper volume spanned by the normalized feature vectors representing di erent textures must be maximized as well. These two criteria provide guidelines for ÿlter bank design.