基于多重插值的亚像素边缘检测

The Detection of Subpixel Edge Based on Multiple Interpolating

亚像素边缘检测是图像处理的重要内容,边缘检测的精度直接影响图像处理的精度。分析一些亚像素边缘检测算法的优缺点,并对基于插值的亚像素边缘检测进行改进。通过粗边缘检测缩小亚像素检测的检测范围,减小计算量。提出多重插值的方法,先对部分点进行插值,将插值结果中的点作为新的插值点,通过两次插值,提高边缘检测的精度和抗干扰能力。从实验结果来看,该亚像素边缘检测算法的最大误差为0.109 pixel,耗时也较短,能应用于实时检测系统。

The subpixel edge detection is an important eontent of image processing, and the precision of edge detection influences the precision of image processing directly. Analyzes the advantages and disadvantages of some subpixel edge detection algorithm and improves the subpixel edge detection of images based on interpolation, reduces the range of deteetion and the amount of calculation by the thick edge detection. Proposes multiple interpolation method, first, calculates part of the interpolation points,then the results of interpolation pninls in as new interpolation points. Through the two limes interpolation, the precision of edge detection and anti-jamming ability are enhanced. From lhe resuhs of experimental, the maximum error of suhpixel edge detection algorithm is 0.109 pixel. Time-eonsuming of the algorithm is shori, so the algorithm can be applied to a real-time detection system.