基于Blackfin561的嵌入式车牌识别系统  

Embedded License Plate Recognition System Based on Blackfin561

作　　者：于鹏[1] 王嘉成[1] 贾伟岗 

机构地区：[1]西安电子科技大学电子工程学院,陕西西安710071

出　　处：《电子科技》2013年第12期135-137,共3页Electronic Science and Technology

摘　　要：介绍了一种基于双核DSP Blackfin561的嵌入式车牌识别系统。讨论了系统的硬件结构和软件实现流程,以及基于Hpofield神经网络进行车牌识别的方法及过程。首先对视频图像进行解码,以获取图像的亮度信号;然后通过图像处理方法对图像进行调整、定位、分割等处理;最终运用Hopfield神经网络识别出车牌上的字符信息。A dual-core DSP Blackfin561 embedded license plate recognition system is introduced. The hard- ware structure and software flow, as well as the neural network-based license plate recognition Hpofield methods and processes are discussed. The f video image is decoded to obtain the brightness of the image signal, then the image is processed such as adjustment, positioning, and segmentation; and finally, Hopfield neural network is used to iden- tify the license plate character information.

关 键 词：车牌识别 Blackfin561 HOPFIELD DSP 神经网络 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]