基于模糊快速Hough变换的盘孔边界识别方法  

Boundary Recognition of an Aperture Based on Fuzzy Fast Hough Transform

作　　者：孙慧贤[1] 张玉华[1] 罗飞路[1] 

机构地区：[1]国防科学技术大学机电工程与自动化学院,长沙410073

出　　处：《光子学报》2010年第2期335-339,共5页Acta Photonica Sinica

基　　金：国防预研重点项目(51317030106)资助

摘　　要：为了快速准确地识别出某型航空发动机篦齿盘均压孔的边界,提出一种新的模糊快速Hough变换.利用局部梯度信息将检测圆所需的3D Hough变换转化为一个2D Hough变换检测圆心位置和一个1D Hough变换检测圆的半径.在检测圆心位置时,对2D参量空间进行多分辨率分级,通过迭代计算由粗到精逐步细化参量空间.采用模糊投票的方法来处理边缘像素位置和梯度方向的不确定性.通过实验对比了模糊快速Hough变换与标准Hough变换、随机Hough变换进行盘孔边界识别的效果,结果表明模糊快速Hough变换可以快速准确地识别出盘孔的边界,性能优于另外两种算法.A new fuzzy fast Hough transform is proposed to detect the circular boundary of Labyrinth disc aperture in an aeroengine. In the new algorithm, the circle detection problem is decomposed into two stages consisting of a 2D Hough transform to find circle centre and a 1D Hough transform to determine the radius. The 2D Hough transform performs a hierarchical approximation to the solution, starting from a coarse description of the parameter space. The coarse;to-fine search technique is used to reduce the computing and storage requirements of the Hough transform..And, the fuzzy vote technique is adopted to lessen the uncertainty that arise from edge pixel position error and gradient direction error. The circular contour detection results by the proposed algorithm are compared with the results by Standard Hough Transform and Randomized Hough Transform respectively. The experimental results show that the fuzzy fast Hough transform can detect the contour quickly and accurately.

关 键 词：计算机视觉 篦齿盘均压孔 边界识别 模糊快速Hough变换 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]