Research And Application Of Wavelet Transform In Image Denoising And Edge Detection

With the development of computer and multimedia technology, image processing technology has been applied to every aspect of life. Image denoising and edge detection is the basis of image processing, so it is necessary to study it. Wavelet transform is a kind of multi scale signal analysis method, and it has the characteristics of low entropy, good time-frequency feature, multi scale analysis and de correlation. These advantages make wavelet transform has a wide range of applications in image denoising and edge detection.It is the first to expound the basic theory of image denoising and edge detection in this paper. These classical methods of image denoising and edge detection are introduced. And the simulation and comparison of these algorithms are carried out to find out the advantages and disadvantages of the classical algorithm, in order to pave the way for future work.Secondly, the basic theory of wavelet transform is studied; it can display both the spatial and frequency domain information of the signal. The size of the wavelet transforms window changes with the change of frequency, which can be used to describe the signal more accurately. In addition, some features of the problem can be highlighted by wavelet transform in order to analysis the time-frequency features of the signal locally.Then, the basic theory of wavelet transform in image denoising is introduced, and the wavelet threshold denoising algorithm is emphatically studied. A new threshold function is proposed by combining the three spline interpolation. And it is better than the traditional threshold function with the advantage of good smoothness. In addition, a new threshold value is constructed by using the artificial bee colony algorithm. it is achieved by the artificial bee colony algorithm to search for profitable signal to noise ratio of the noisy image. The experimental results show that this algorithm is effective.At last, the application of wavelet transform in edge detection is introduced. And the methods of edge detection with wavelet transform modulus maxima are emphatically studied. A new edge detection algorithm is proposed based on the knowledge of omni-directional wavelet transform, morphology and Hausdorrf distance. It set the detected edge with omni-directional wavelet transform as A; and set the edge of improved morphological detection as B. In the end, the distance of the A and B is calculated by the improved Hausdorrf distance, and the distance is taken as the finally edge point. The experimental results show that this algorithm is effective.