This paper presents an architecture for the trajectory tracking of a Khepera II wheeled mobile robot in a certain environment, which means that the workspace of the robot is supervised by a camera, placed vertically above. The proposed strategy consists of computing the position of the robot in real time using a vision algorithm. Then the technique for tracking-error model-based Parallel Distributed Compensation (PDC) control for non-holonomic vehicles is used. Briefly, this technique consists of rewriting the kinematic error model of the mobile robot tracking problem into a Takagi-Sugeno (TS) fuzzy representation and finding a stabilizing controller by solving Linear Matrix Inequalities (LMI) conditions for the tracking-error. The effectiveness, perfect trajectory tracking and real time properties are shown by experimental results.