Design and Implementation of Driver Training Expert System based on Driving Posture Recognition

This paper presents the design and implementation of driver training expert system which is created to coach and train the person especially driving learner who lacks of the correct driving posture in various sequence of driving postures. Webcams are used as visual sensors to obtain the driving leaner posture as input parameters. Image processing, Viola-Jones and Skin Color Segmentation algorithms are used to recognize driving learner postures. A rule-based expert system is built to evaluate whether the detected driving posture is correct by considering the relation between head movement with the activities on steering wheel and shift lever. If incorrect driving posture has occurred, this proposed method can be corrected it by giving the proper advice. The experimental shows the effectiveness of the proposed method in coaching and training the driving learner.