摘要
A mixed scheme based on Wavelet Transformation (WT) is proposed for image edge detection. The scheme combines the wavelet transform and traditional Sobel and LoG (Laplacian of Gaussian) operator edge-detection algorithms. The precise theory analysis is given to show that the wavelet transformation has an advantage for signal processing. Simulation results show that the new scheme is better than only using the Sobel or LoG methods. Complexity analysis is also given and the conclusion is acceptable, therefore the proposed scheme is effective for edge detection.
A mixed scheme based on Wavelet Transformation(WT) is proposed for image edge detection. The scheme combines the wavelet transform and traditional Sobel and LoG(Laplacian of Gaussian) operator edge-detection algorithms. The precise theory analysis is given to show that the wavelet transformation has an advantage for signal processing. Simulation results show that the new scheme is better than only using the Sobel or LoG methods. Complexity analysis is also given and the conclusion is acceptable, therefore the proposed scheme is effective for edge detection.
基金
Supported by the National Defence 973 project(2002HS0604,2002HS0634)
关键词
波长变换
边缘检测
多分辨分析
正交性
图像处理
Wavelet Transformation(WT)
Edge detection
Multiresolution analysis
Orthogonality