Application Research Of Dyadic Wavelet Transform In Image Edge Processing

Image edge detection is one of the important and difficult technical problems inthe research of many scholars at home and abroad, is the most active research topicsin the field of image processing and various domains, is very important basis of theimage future detection. The primary job in the image processing is edge detection,plays a very big promoting role to the development of image processing and othervarious areas, is one of the most promising image processing method. In recentdecades, the basic theory of wavelet transform has also been a continuousimprovement, the wavelet transform is the most popular in the field of image edgedetection. Discrete wavelet transform as a most promising image processingtechnology, has received many researchers in the different areas. Among them, themost suitable for our various needs is bi-orthogonal wavelet transform and dyadicwavelet transform. Discrete wavelet transform as a new effective image processingtools, have been subject to the attention of a lot of people. Among them, the mostsuitable for our wide variety of needs is the bi-orthogonal wavelet transform anddyadic wavelet transform. Both transform in image processing have their respectiveadvantages and disadvantages, bi-orthogonal wavelet transform in image processingwith non-translational invariance and down sampling, and dyadic wavelet transformin image processing with translational invariance and without down sampling. Fromthe wavelet transform theory appears to now, we have the dyadic wavelet transformimage edge detection method is not satisfactory. In order to achieve the rapiddevelopment of the field of edge detection, need to improve the dyadic wavelettransform image edge detection technique, seriously study the traditional image edgedetection operator, and comparative analysis of dyadic wavelet transform, on dyadicwavelet transform in image edge detection conduct a comprehensive study.Now has a traditional edge detection methods are many inherent flaws, forexample, although the traditional image edge operators easy to implement, can getbetter results in a noise-free image, but to the image of lower signal-to-noise ratiothere are details missing serious, poor ability of anti-noise, inaccurate positioning etc.And because of the inherent complexity of image edge detection problems, theseoperators often can not get satisfactory results. So need to deeper study the edgedetection applications of dyadic wavelet transform. Through research, on the basis ofthe existing image edge detection methods, mainly applied dyadic wavelet transform,to dyadic wavelet transform combined application the histogram equalization imageenhancement method and wavelet image fusion, presented several methods of image edge detection based on dyadic wavelet transform. Conduct, compare the detectioneffect of traditional image edge detection method and presented method by Matlabsimulation experiments, also do some theoretical research. Simulation results showthat, in the terms of noise immunity, edge detection, detection accuracy, the presentedmethod is much better than directly using traditional methods or dyadic wavelettransform image edge detection method to the original image, is more effective andgood application prospects image edge detection method.