压缩感知在车牌识别系统中的应用研究

分析车牌识别系统的数据采集过程,提出关于压缩感知理论的车牌识别系统的数据采集方法。对采集的车牌图像进行小波变换,将车牌图像变换成易传输,数据量较小的图像数据,使用正交匹配追踪算法以及小波逆变换对来自传输设备的车牌图像进行重建,并实现原始图像精确和近似重构。由仿真可知,压缩感知理论应用于车牌识别系统的数据采集传输过程,降低了传输数据量大小,并易于图像数据传输,且能以较小误差实现车牌图像的重构,同时有相应的硬件使用环境。

Compressed Sensing and its application in license plate Recognition System

The data collection process in license plate system are analyzed.Data collection method is proposed based on the compressed sensing license plate recognition system.Wavelet transform for acquired image,converted the license plate iamge into image data which is easy transfer and small,use OMP(orthogonal matching pursuit) algorithm and inverse wavelet transform reconstruct the license plate image from transmission equipment,and to achieve the precise and approximate the image reconstruction.From the experiment and the simulation calculation,compressed sensing theory is applied to data collection and transmission process in license plate recognition system,reducing the amount of data transfer size,to facilitate the image data transmission,and can achieve license plate image reconstruction by a smaller deviation,adapting to corresponding hardware environment.