Abstract : This thesis deals with the mobile robot self_guidance detection, recognition and tracking, which heads towards a classified object through a combination of hardware and software processing, where an object is selected through a live screen selection then detected through several methods in addition to state predication of the next position concerning mobile robot instant tracking . The mobile robot will head towards the specified object, finally stops before touching it. The self_guidance mobile vehicle application is provided with an implementation of hardware interface circuit for the computer and the vehicle along with software implementation of image processing system design, all proposed in this thesis text. The software application is programmed using Borland C++ Builder v.6 Professional, windows compatibility. Mobile robot is supported with user real-time GUI (Graphical User Interface) environment, which is used for object selection, tracking methods activation and switching, and also for extraction of the general gained information in addition to vehicle control, with minimum processing delay. Mobile Robot software implements, _snake_ active contours detection combined with threshold filtering, an advance technique of histogram recognition, besides motion detection and tracking application, obstacle detection and finally it implements non-extended linear kalman filter for state prediction. The goal is to achieve minimum delay of processing with maximum performance, to enhance future researches in this field in order to create home made robot, with optimum applications. The model of mobile robot presented was successfully able to detect, identify and track (a fixed object, a moving object) in both ideal and non-ideal environment, once the original object_s recognition parameters stored in memory.