A Novel Method for License Plate Location

Jiawen Song,Chuang Zhu,Xiaodong Xie,Yuan Li,Meng Wang,Huizhu Jia

DOI: https://doi.org/10.2991/amcce-18.2018.133

2018-01-01

Abstract:license plate location is one of the key steps for license plate recognition. With license plate recognition technology, police can better manage vehicle and this technique also give convenience to application such as parking fee charge. In this paper, we propose a novel method to locate license plate of china in complicate environment. First, we change the RGB image into greyscale image. Then, we apply Sobel operator on the grey-scale image to get its Sobel edge image. After that the Sobel edge image is change into a binary image in which each pixels of the image are decided by statistic its variance in nearby vertical and horizontal pixels. Base on the binary image above, the algorithm locates the candidate boxes and produces a list for them. Next, we select the boxes above by using hsv colour space and then merge the boxes base on their location. At last, the unreasonable boxes are removed and the final location of license plates produced. Our method has outperformed two types of license plate location products in a small surveillance dataset and achieved a fast and high precision location result compared with the-state-of-art.