A LICENSE PLATE DETECTION METHOD BASED ON MORPHOLOGY AND MAXIMALLY STABLE EXTREMAL REGION

This paper presents a novel license plate extraction algorithm based on the combination of morphology and Maximally Stable Extremal Region. The method can achieve a high accuracy and efficiency on the real scene images. The proposed approach is mainly composed of four steps: image enhancement by the top-hat transform, MSER detection with aspect ratio threshold, candidate plate regions extraction, and license plate verification. Under the experiment databases of 8765 real scene images, the average accuracy of locating license plate is 98.9%.Computer Science, Theory & MethodsEngineering, Electrical & ElectronicCPCI-S(ISTP)