| NSCT的局部能量算法在图像特征检测中的研究 |
| |
| 引用本文： | 马容生.NSCT的局部能量算法在图像特征检测中的研究[J].柳州职业技术学院学报,2010,10(4):57-61. |
| |
| 作者姓名： | 马容生 |
| |
| 作者单位： | 柳州职业技术学院信息工程系,广西柳州545006 |
| |
| 摘 要： | 图像特征检测是图像处理的重要步骤,也是图像处理的重点和难点。大部分特征检测算法都是基于图像灰度幅值之间的差异,本文使用非下采样变换（NSCT）计算图像的局部能量从而检测图像的构建特征。这种算法可以检测多种图像特征,特别是在检测线性特征方面具有优势。由于NSCT没有使用下采样,这种算法具有平移不变性,更有利于对多幅图像进行处理。从试验结果可以看出相对于经典的canny算法和基于Gabor的算法,NSCT效果更好。
|
| 关 键 词： | 局部能量 NSCT变换 多尺度 特征检测 |
A Study of Local Energy Detecting Image Features Based on NSCT |
| |
| MA Rong-sheng.A Study of Local Energy Detecting Image Features Based on NSCT[J].Journal of Liuzhou Vocational & Technical College,2010,10(4):57-61. |
| |
| Authors: | MA Rong-sheng |
| |
| Institution: | MA Rong-sheng(Information Engineering Dep.,Liuzhou Vocational ＆ Technical College,Liuzhou 545006,China) |
| |
| Abstract: | The detection of image features is the most important step in image process,which is also the most difficult step.Most feature detection algorithm is based on the amplitude of the difference between the gray levels in an image.In the paper we use the Non-subsampled Contourlet Transform（NSCT）to calculate the local energy to extract image built features.This method can be used to detection kinds of image features and do well in line feature.Without using subsample,NSCT has the translation invariance.This is more beneficial to processing multiple images.Experimental results show the proposed method achieves better than classical canny algorithm and the algorithm based on Gabor. |
| |
| Keywords: | local energy NSCT multi-scale feature detection |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
