No abstract

Abstract-Analysing micro-and macro-structures within images confers ability to include scale in texture analysis. Filtering allows for selection of texture structures at different scales, revealing the micro-and macro-structures which would otherwise be concealed. The new approach to texture segmentation uses low-and high-pass filters to achieve this scale-based analysis. Segmentation is performed using Local Binary Patterns as an example of the type of feature vector that can be used with the new process. These are generated for the original image and each of the filtered images. A two stage training process is used to learn the optimum filter sizes and to produce model histograms for each known texture class. These are used in the supervised segmentation of texture mosaics generated from the VisTex database. The results demonstrate the superiority of the new combined approach compared to the best multi-resolution LBP configuration and analysis only using lowpass filters. Noise analysis has also confirmed the advantageous properties of low-and high-pass filtering, and confirms that it is optimal to combine the two forms in texture segmentation.

A new approach to colour-texture segmentation is presented which uses Local Binary Pattern data and a new colour quantisation scheme based on hue and saturation to provide evidence from which pixels can be classified into texture classes. The proposed algorithm, which we contend to be the first use of evidence gathering in the field of texture classification, uses Generalised Hough Transform style R-tables as unique descriptors for each texture class. Tests on remotely sensed images demonstrate the superiority of the colour-texture algorithm compared to the established JSEG algorithm; a notable advantage of the new approach is the absence of over-segmentation. The VisTex database is used to compare the colour-texture algorithm with alternative methods, including its grey-scale equivalent, for the segmentation of colour texture images; providing good results with smooth texture boundaries and low noise within texture segments.