摘要
图像测量需要高精度、高准确度的图像边缘。随着测量精度要求提高,像素级图像边缘不能满足实际的需求,需要更为精确,提供更多细节的亚像素级边缘。提出了一种在HSV彩色空间利用基样条插值提取亚像素边缘的算法。首先,将RGB空间的彩色图像转换到HSV颜色空间;然后,利用提出的HSVsin-cos彩色距离模型检测出像素级的边缘;最后,对像素级边缘进行基样条插值,得到亚像素边缘,并与B样条、三次样条插值结果相比较。实验结果表明:基样条插值的结果对图像的逼近度、差异性优于其它方法,分析与讨论的结果为利用插值方法进行亚像素边缘精确定位技术选取提供可靠的依据。
Image measurement prefer higher image edges precision.Sub-pixel edges which offer more details and higher precision were explored instead of traditional pixel-level image edges as the required precision got increasingly higher.It took three steps for this article to propose analgorithm which intends to extract sub-pixel edge using Based-Spline in the HSV color space.Firstly we transferred the color space to color space;then,using the extracted HSVsin-coscolor range model we checked out pixel edges;and finally,through based-spline interpolation on pixel-level edge,we got subpixel edge and made comparison of the splines and cubic spline interpolation results.Experimental results show that the approximation and difference of the base spline interpolation results are superior to the cubic spline interpolation methods,analysis and discussion provides a reliable basis of the results using the interpolation method for sub-pixel edge precise positioning technology.
出处
《计算机与数字工程》
2015年第6期1107-1112,共6页
Computer & Digital Engineering
基金
陕西省自然科学基础研究计划项目(编号:2013JM8043)资助
关键词
边缘检测
亚像素
彩色图像
样条插值
信噪比
edge detection, sub-pixel, color image, spline interpolation, SNR
