基于ridgelet变换的图像直线特征检测算法研究  

Research on Image Straight Line Feature Detection Algorithm Based on Ridgelet Transform

作　　者：罗扬[1] 薛俊玲[1] LUO Yang;XUE Jun-ling(Jinling Institute of Technology, Nanjing 211169, China)

机构地区：[1]金陵科技学院软件工程学院,江苏南京211169

出　　处：《金陵科技学院学报》2021年第2期19-24,共6页Journal of Jinling Institute of Technology

基　　金：江苏省现代教育技术研究课题(2021-R-92099);江苏省高等学校自然科学研究重大项目(19KJA510004)。

摘　　要：常用的图像直线特征检测算法如霍夫变换(Hough transform)和拉东变换(Radon transform),存在计算量大、抗噪能力弱等问题。为此,提出了基于脊波变换(ridgelet transform)的图像直线特征检测算法,其作为一种多尺度几何分析与稀疏表示的应用算法,具有良好的区域平滑性、可逆性和去冗余性。实验结果表明,在复杂噪声干扰的环境下,基于ridgelet变换的图像直线特征检测算法能够获得清晰的直线检测结果,具有较好的鲁棒性。The commonly used image straight line feature detection algorithms,such as Hough transform and Radon transform,have the problems of large amount of calculation and weak anti noise ability.Therefore,an image straight line feature detection algorithm based on ridgelet transform is proposed.As an application algorithm of multi-scale geometric analysis and sparse representation,it has good regional smoothness,reversibility and de redundancy.The experimental results show that the image straight line feature detection algorithm based on ridgelet transform can obtain clear line detection results and has good robustness in the complex noise environment.

关 键 词：直线特征检测算法 HOUGH变换 RADON变换 RIDGELET变换 图像 

分 类 号：TN919.81[电子电信—通信与信息系统]