非均匀光照下的车牌图像分割算法

Segmentation algorithm for unevenly illumination license plate image

为了克服非均匀光照对车牌图像分割效果的影响,提出了一种新的光度补偿插值算法,将子图像的平均灰度值作为中心相对照度,并且利用插值算法得到子图像边缘照度,将非均匀光照的车牌图像转变成均匀光照图像,进而利用Otsu算法对均匀光照车牌图像进行分割。实验结果表明,该算法对非均匀光照图像分割精度高,运算速度快,具有更加理想的分割效果。

Aiming at overcoming the influenceof unevenly illumination to the license plate image segmentation,this paper proposes a new illumination compensation interpolation algorithm. The average gray value is regarded as the relative intensity at center of sub image,and using the interpolation algorithm to get the edge illumination image.This algorithm converts unevenly lighted license plate image to evenly image. Then through Otsu algorithm,it obtains the binary license plate image. The results show thatthe algorithm has segmentation of high precision and obtains more perfect image segmentation results.