Texture Segmentation Using Gabor Filters

The human’s capability to distinguish perceptually different textures is difficult to reproduce using machine vision due the variety of textural patterns and illumination conditions. Bovik et al. proposed a multi-channel texture analysis technique that relies on 2-D Gabor filters to isolate regions of perceptually homogeneous texture in an image. Textures are modeled as a pattern dominated by a narrow band of spatial frequencies and orientations. Properly tuned Gabor filters react strongly to specific textures and weakly to all others. The amplitude of the channel outputs is compared to segment the textures. The phase of the channel outputs is used to locate discontinuities in the texture phase. In this paper the theory proposed by Bovik et Al. is revisited and the results of experimentation with the technique are presented. Texture segmentation using the magnitude of the channel response is shown to be robust with a variety of synthetic and real textures. Experimentation with texture segmentation using the phase of channel response is inconclusive.