运用压缩传感理论的交通视频图像处理

Using compressive sensing theory in traffic video image processing

目前图像的编解码都是运用传统的香农定理进行的,此定理要求采样频率要大于等于原始信号最高频率的两倍。在进行视频图像处理时,就产生了需要相当大的存储空间来存储获得的数据的问题。鉴于此,提出了运用压缩传感理论进行编解码的方法,它可在相同的压缩率下产生更少的压缩数据,进而可以节约存储空间和传输时间。首先进行了传统的信号获取和处理与压缩传感理论的比较,再对压缩传感理论进行了阐述,最后把压缩传感理论应用于交通视频图像处理中,获得了较好的效果,这也是压缩传感理论的一次尝试性的应用。

At present,image encoding and decoding is used with the traditional shannon theorems which requires that sampling frequency must be twice or more than highest frequency of original signal.In video image processing,it produces the problem that it need quite a greater storage space to store the data obtained.For this reason,It is proposed the method which uses compressed sensing theory to encode and decode,With the same compressive rate,it can produce less compressive data which would save more memory space and transmission time.Firstly,this paper compares the traditional signal acquisition and processing,then states the CS theory,at the last,it applys CS theory in the processing of the traffic video image and obtains good result,which is also a tentative application of CS theory.