There has been a great deal of research efforts trying to decide the proper action of a robot performing a task in an unknown environment by using vision information. Specially, One of the main problems in feature-based visual servoing is difficulties involved in mathematical modelling, acquisition of range data and uncertainty of vision information. In this thesis, an image feature-based fuzzy self-organizing visual tracking controller is proposed to solve the problems mentioned above. The proposed algorithm uses uncertain image feature without range data for the control of eye-in-hand robot. Several experimental results show the usefulness of the proposed algorithm.