统计纹理分析寻找刻划纹理的数字特征，用这些特征或同时结合其他非纹理特征

Statistical texture analysis finds digital features that characterize textures, and uses these features or other non-texture features to classify regions (not individual pixels) in the image. The autocorrelation function, gray-level co-occurrence matrix, gray-scale run-length and various statistics of gray-scale distribution of image local area are commonly used digital texture features. For example, the gray level co-occurrence matrix uses the spatial distribution of gray levels to characterize the texture. Because the gray distribution of coarse texture changes much more slowly than fine texture, the two have completely different gray co-occurrence matrices.

Statistical texture analysis looks for the digital features of the engraved texture and categorizes the areas in the image instead of a single pixel with these features or in combination with other non-texture dilike features. The self-correlation function of the local area of the image, the grayscale symbiosis matrix, the grayscale travel, and the various statistics of grayscale distribution are commonly used digital texture features. Such as grayscale symbiotic matrix with grayscale spatial distribution characterizes the texture. Because the grayscale distribution of coarse textures varies much more slowly with distance than fine textures, the two have completely different grayscale symbiosis.

Statistical texture analysis is used to find the digital features that describe the texture, and these features or other non texture features are used to classify the regions in the image (rather than a single pixel). The autocorrelation function, gray level co-occurrence matrix, gray level run and various statistics of gray level distribution in local area of image are commonly used digital texture features. For example, the gray level co-occurrence matrix uses the spatial distribution of gray level to represent the texture. Because the gray distribution of coarse texture changes more slowly with distance than that of fine texture, they have totally different gray level co-occurrence matrix.<br>