利用图像分解和Beamlet变换的含噪图像线特征检测

Noisy Image Line Feature Detection Method via Image Decomposition and Beamlet Transform

本文提出了一种含噪图像线特征检测算法。根据图像分解理论和Beamlet变换的特点,将含噪图像的结构和纹理进行分量进行分解,以降低噪声和纹理对线特征检测的影响,再利用Beamlet变换对图像进行线特征检测。通过对富含线特征和边缘纹理复杂的含噪图像进行实验,与Donoho的原检测算法进行了比较。实验结果表明本文算法的有效性。

This paper proposes a linear feature detection method for noisy images.According to image decomposition theory and the characteristics of beamlet transform,the structure and texture components of noisy image can be decomposed,to reduce the influence of noise and texture on line feature detection.Then the beamlet transform is applied to extract the linear feature.Compared with Donoho＇s method,the experiments on line-rich and complex edge natural images show that the proposed algorithm can achieve satisfactory performance in the cases of low signal-to-noise ratios.