The Co-occurrence Frequency Image: A New Texture Feature Image Paradigm

Image processing is generally based on grayscale information, such as statistical gray histograms, gradients, textures, etc. However, concrete techniques that use this co-occurrence processing have been chiefly limited to texture identification. Since the gray histogram is the most influential for extracting the global image properties, it has always been desirable to improve or extend it. From this viewpoint, Kashiwagi's proposal of a new frequency extraction paradigm called the frequency image (FI) is noteworthy. Along with other histograms, the co-occurrence histogram or co-occurrence matrix is also an influential basic image feature, especially for texture feature analysis. In this paper we introduce a new texture feature image paradigm called the "co-occurrence frequency image" (CFI), which is a complete extension of the FI. (C) 2010 Wiley Periodicals, Inc. Electron Comm Jpn, 93(12): 50-59, 2010; Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ecj.10145