摘要
利用超分辨率重建技术提高车牌图像的质量,以获取更好的识别效果。首先采用基于小波变换的块匹配方法对车牌图像进行配准,然后利用迭代反投影算法对配准后的图像进行重建获得高分辨率的图像。对模拟生成的多帧图像和标准测试视频序列进行了实验,实验结果表明:该算法能有效提高车牌图像的质量。
In order to improve the quality of a vehicle license plate image, the super-resolution reconstruction technology is used.Firstly, block matching method based on wavelet transform is applied to obtain the registration of the image.And then, iterative back projection algorithm is used to reconstruct a super-resolution image from the registered image.Reconstruction is performed on simulated image and standard test video respectively. Experimental results show that the algorithm can effectively improve the quality of a vehicle license plate image.
出处
《成都工业学院学报》
2015年第4期18-21,共4页
Journal of Chengdu Technological University
基金
四川省教育厅科研项目"基于目标分割的超分辨率重建研究"(14ZA0283)
关键词
超分辨率
图像重建
块匹配
迭代反投影
super-resolution
image reconstruction
block matching
iterative back projection
