Detection Algorithm based on Wavelet Threshold Denoising and Mathematical Morphology

摘要

In this paper, the threshold function denoising algorithm and mathematical morphology are combined and applied to image edge detection. Firstly, we construct a dyadic wavelet with non-orthogonality, symmetry, limited spectrum, and smoothness almost everywhere. Then, the properties of the dyadic wavelet are discussed, and the analytic expression of the reproducing kernel function in the image space of dyadic wavelet transform is given. Moreover, the dyadic wavelet is used to construct a new threshold function for image denoising, and the new threshold function has a clear effect on image denoising. Finally, we present an improved morphological edge detection algorithm, which is applied to extract the edges of images after threshold denoising. Thus, we can obtain a new edge detection algorithm that combines the threshold function denoising algorithm and morphological edge extraction algorithm. The simulation results show that the edges detected by the new algorithm are clearer and contain less noise, and the continuity and accuracy are also improved.