MVSR Normalization Algorithm Method for Improving Vehicle License Plate Recognition

Abstract

Image processing and embedded systems are used in many applications such as object recognition, robotic applications, autonomous and remote control systems developed for the defense industry, medical applications, face recognition, and vehicle license plate recognition. Many vehicle license number plate detection methods are not effective under vehicle license plate images have a degree of rotation or low resolution images. Thus, we used MVSR normalization algorithm to detect vehicle license plate recognition for better accuracy and lower computational cost. The MVSR normalization algorithm, Mean–Variance-Softmax-Rescale processes respectively is applied for high-accuracy real-time vehicle license plate detection.