摘要
论文针对实时交通路况图像,定义了两种描述图像扫描数据抖动性的度量指标:局部性指标和连续性指标。证明了Hilbert曲线扫描优于其他传统扫描;分析了图像扫描方式对基于多项式拟合图像压缩编码效果的影响;实验还表明,这两种度量方法不仅可以对现有图像扫描方法进行评价,还可以指导设计新的图像扫描方法。基于多项式拟合的图像压缩方法是该文的另外一个新思路。实验表明,它简单且有效。结合光栅扫描的基于多项式拟合方法对象素为256乘以256的复杂交通图像可以得到峰值信噪比大于20且压缩比小于20%的恢复图像。
This paper defines two kinds of measurement parameter to describe the stability of the image scanning data for the online traffic image.The two measurement parameters are Local-parameter and Continuity-parameter.The experiments have compared five kind of scanning ways,the raster scanning,the helix scans,the Z scanning,the sequence raster scanning and the Hilb＇ert scanning.Experiments have analyzed the image scanning＇ affect to the result of image compression based on polynomial approaching.Experiments prove Hilbert curve scanning are steadier than other scanning ways.The measurements can not only evaluate the known method of image scanning,but also guide the design for the method of image scanning.Image compression based on polynomial approaching is another innovation,The method with raster scanning is simple but effective.The Power Signal-to-Noise Ratio is bigger than 20,while the compression rate is smaller than 20% to the small complex traffic image which is 256 multiply 256.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第25期217-219,共3页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(编号:60134010)
关键词
图像扫描
局部性指标
连续性指标
交通图像压缩
image scanning,local parameter,continuity parameter,traffic image compression
