Muscle fiber types can be described using histochemical, biochemical, morphological, or physiologic characteristics; however, classifications of muscle fibers by different techniques are not always agreed. Therefore, muscle fibers that may be grouped together by one classification technique may be placed in different categories using other techniques. A basic understanding of muscle structure and physiology is necessary to understand the muscle fiber classification techniques. Histochemical staining of skeletal muscle specimens with alpha naphthyl acetate esterase (ANAE) reaction is considered as the main step of the input image acquisition of our system, by using a microscope and a photograph. Fuzzy clustering algorithm is used in this thesis to analyze of biopsy images of muscle tissue with different fiber types. Segmenting of the image is achieved by a system composed of two sequential stages as follows: First stage includes implementation of unsupervised fuzzy clustering algorithm in order to outline the regions in the input muscle biopsy image. Extracted regions are spatially localized and have uniform brightness. The second stage involves separation of different types of the muscle fibers. It consists of three steps and is based on a threshold technique. At this stage, different types of the human muscle fibers are assigned into one of the following regions: slow-twitch oxidative fibers (Type I), fast-twitch oxidative fibers (Type IIA or one of its subtypes) and fast-twitch glycolytic fibers Type IIB, or the endomysium that might be recognized in the image. In addition, the effects of the histochemical reaction time and the special optical filters on the image are studied. The system is implemented using MATLAB 6.5.