Image processing such as Traffic Sign Recognition (TSR) plays a key role in Intelligent Transportation Systems particularly in Traffic Sign Recognition (TSR) which aims at increasing driver safety. Several studies have proposed TSR systems based on different image processing and machine learning algorithms. However, the efficiency of the proposed TSR algorithms requires improvement to enable real-time alerts on onboard devices which have limited computational power. Further improvement on accuracy of TSR is also required mainly in unstable weather conditions or when multiple signs exist on one pillar. This research proposes an improved model for Automatic TSR (ATSR) consisting of improved Connected Components Labeling and color histogram. Firstly, images are captured in real-world conditions in Palestine. The region of interests is detected using the improved Connected Components Labeling algorithm combined with Vectorization to reduce computational time. Secondly, the local features of detected images are computed using the color histogram. The results of the research show that combining the Connected Components Labeling with Vectorization reduces the computational time. Also, the Connected Components Labeling algorithm followed by histogram increase the accuracy of recognition. The low computational cost and the accuracy of the model enable us to use the model on smart phones for accurately recognizing traffic signs and alerting drivers in real time.