基于二元向量值有理插值的彩色图像重建  

COLOUR IMAGE RECONSTRUCTION BASED ON BIVARIATE VECTOR VALUED RATIONAL INTERPOLATION

作　　者：黄日朋[1] 王大星[1] 

机构地区：[1]滁州学院数学科学学院,安徽滁州239000

出　　处：《计算机应用与软件》2013年第5期155-158,187,共5页Computer Applications and Software

基　　金：安徽高校省级自然科学研究项目(kj2011z285);滁州学院科研基金项目(2010kj008B)

摘　　要：彩色图像比灰度图像含有更多的信息,描述彩色图像特征的比特数远大于灰度图像,因此数据压缩在彩色图像处理中就起到了重要的作用。传统的方法是将彩色图像分解成多通道的单色图像处理,但这样容易造成图像颜色信息的丢失。考虑到这一缺点,提出一种新的图像重建方法,即在RGB空间中通过对原图像进行采样,构造一种二元向量值有理插值函数,计算采样点之间数据点的像素值,并实现彩色图像的重建。实验结果表明只需原图的30%或更少的像素值即可重建原图像,且重建后的图像边缘清晰,细节保持较好,平均误差率较小。Colour image contains more information than the grayscale image has,and its bits number describing the characteristics is much larger than the grayscale image＇s as well.Therefore,data compression plays an important role in colour image processing.Traditional method is to decompose the colour image into monochrome images of multi-channel for processing,but it is likely to lose colour information of the image.Concerning this shortcoming,we propose a new image reconstruction method.It samples the original image in the RGB space and constructs a kind of bivariate vector valued rational interpolation function to calculate the pixel values of data points between the sampling points,and achieves reconstruction of the colour image.Experimental results show that it just needs 30% or fewer pixel values of the original image to reconstruct the original image,and the reconstructed one maintains sharp edges,precise details and relatively small average error rate.

关 键 词：彩色图像重建 向量值有理插值 向量值梯度算子 彩色边缘提取 RGB空间 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]