A License Plate Detection Algorithm Based on Deformable Part Model

摘要

This paper proposes an algorithm for license plate detection based on deformable part model, as it is difficult to obtain a high accuracy in detecting license plates in images with complex background and loose perspective. We have researched the state-of-the-art object detection method making use of the deformable part model, combined its training method with semi-supervised learning SVM, and introduced the method to the problem of license plate detection. The result shows the validity of the proposed method.