In computer vision, Texture segmentation is the process of partitioning a digital image into multiple texture segments.

The goal of texture segmentation is to simplify or change the representation of an image into something that is more meaningful and easier to analyze. Texture segmentation is typically used in SAR images to locate regions and boundaries (lines, curves, etc.) in images. More precisely, texture segmentation is the process of assigning a label to every pixel in an image such that pixels with the same label share certain visual characteristics. The result of texture segmentation is a set of texture segments that collectively cover the entire image, or a set of contours extracted from the image (see edge detection). Each of the pixels in a region are similar with respect to some characteristic or computed property, intensity, or texture. Adjacent regions are significantly different with respect to the same characteristic(s).

Introduction
The classification of land use from high resolution satellite imagery is generally regarded as a difficult task with much depending on the skill of the analyst, quality and accuracy of the data, and the precise classification methodology. Even more difficult is the classification of those land cover types that make up the broader class of urban land use. Buildings and roads tend to get lumped together along with parking lots and driveways as impervious areas. This general land cover/use category is a heterogeneous mixture of asphalt roads, concrete, steel, gravel and other dissimilar materials. They all have similar spectral signatures when contrasted against the surrounding ground covers (i.e. trees, bushes, and other foliage). This makes it difficult to extract a specific land cover from this mixture of spectrally similar land-covers using pixel based classification methods that rely solely on the average spectral reflectance of a given area. This is further limited by the spatial resolution of the instrument used to record said reflectance. LBP was first mentioned by Harwood et al., and was introduced to the public by Ojala et al. They described the comparative change of gray values between the samples in the circular local neighborhood as texture structural features and made a statistical analysis for these texture features using texture spectrum histogram. Texture information of images, however, can not be described adequately only using LBP operator with small radius because therere also some big structural information in the images. As a consequence, it is necessary to use multi-resolution versions. Ojala developed original LBP and made it a more flexible operator which is circular form with discretionary radius and neighborhoods. This makes it possible to acquire different resolution LBP operators. The LBP texture analysis operator is defined as a grayscale invariant texture measure, derived from a general definition of texture in a local neighborhood. For each pixel in an image, a binary code is produced by thresholding its value with the value of the center pixel. A histogram is created to collect up the occurrences of different binary patterns. The basic version of the LBP operator considers only the eight neighbors of a pixel, but the definition has been extended to include all circular neighborhoods with any number of pixels. Through its extensions, the LBP operator has been made into a really powerful measure of image texture, showing excellent results in terms of accuracy and computational complexity in many empirical studies.

The LBP operator can be seen as a unifying approach to the traditionally divergent statistical and structural models of texture analysis. Perhaps the most important property of the LBP operator in real world applications is its tolerance against illumination changes. Another equally important is its computational simplicity, which makes it possible to analyze images in challenging real-time settings. The LBP method has already been used in a large number of applications all over the world, including visual inspection, image retrieval, remote sensing, biomedical image analysis, face image analysis, motion analysis, environment modeling, and outdoor scene analysis. My work considers the Remote Sensing applications involving High Resolution Satellite Imagery.

Objectives: (i) (ii) Study of various existing segmentation technique. Advantage and disadvantage of using this technique on high resolution images. Development of suitable software for segmenting high resolution images.

Methodology
It has been planned to develop the basic LBP operators and test them on satellite images. The output would be tested and compared to the output of the initial segmentation. Thereafter an operator based on combination of the circular LBP and Variance Operator is to be developed and the output to be tested as done before. Thereafter a new operator based on geometrical spatial combinations of pixels like diagonals etc would be developed and tested as before. The proposed method is expected to give better segmentation results as it would consider a large no of determining pixel intensities.