Autor/Autorin: Sager Kamal
Kamal H. Sager, Ph.D.: Studied Astronomy and Space Sciences/Digital Image Processing at Baghdad University, Had a Post-Doc in Biomedical Imaging Researches from Tokyo Institute of Technology, JAPAN. Assistant Professor at College of Pharmacy,Mustansiriyah University, Baghdad, IRAQ.
Color texture classification is an important step in image analysis specially for segmentation and recognition applications of textures of natural scenes, such as leaves surfaces, terrains models, etc. and also in medical applications like hist-Pathological tissue analysis. This book introduce the idea of using the concept of fractal dimension spectrum for describing, discriminating and classifying different visual textures. Two methods for estimating the fractal dimension spectra are proposed; the first method is the modified binarization method and the second is the traditional box counting method. A testing procedure was conducted to evaluate the degree of sensitivity for each kind of fractal dimension spectrum to the visual variation of the different textures. It has been shown a satisfactory results for the usage of fractal dimension spectra as textural discriminating criteria. The proposed classification methods were applied on two important set of images; the first set consists of remote sensing images, and the second set consists of breast tumors images. The higher recognition accuracy in the first set of textures images was (100%), while in the second set was (98.8%).
Über den Autor
Kamal H. Sager, Ph.D.: Studied Astronomy and Space Sciences/Digital Image Processing at Baghdad University, Had a Post-Doc in Biomedical Imaging Researches from Tokyo Institute of Technology, JAPAN. Assistant Professor at College of Pharmacy,Mustansiriyah University, Baghdad, IRAQ.
Klappentext
Color texture classification is an important step in image analysis specially for segmentation and recognition applications of textures of natural scenes, such as leaves surfaces, terrains models, etc. and also in medical applications like hist-Pathological tissue analysis. This book introduce the idea of using the concept of fractal dimension spectrum for describing, discriminating and classifying different visual textures. Two methods for estimating the fractal dimension spectra are proposed; the first method is the modified binarization method and the second is the traditional box counting method. A testing procedure was conducted to evaluate the degree of sensitivity for each kind of fractal dimension spectrum to the visual variation of the different textures. It has been shown a satisfactory results for the usage of fractal dimension spectra as textural discriminating criteria. The proposed classification methods were applied on two important set of images; the first set consists of remote sensing images, and the second set consists of breast tumors images. The higher recognition accuracy in the first set of textures images was (100%), while in the second set was (98.8%).
