基于小波图像分割技术研究.doc

基于小波的图像分割方法研究 摘要 在过去的几十年，计算机系统的处理速度，存储能力以及计算机网络都发展着，这些的发展数字图像处理以及网络多媒体提供了有力的技术保障，并且大的了数字图像处理及网络多媒体的研究和应用。图像处理已经成为一个的学科，其研究领域和研究课题数不胜数。数字图像处理理论和技术大致可概括为:图像数字化、图像、图像、图像、图像、图像分割、特征提取等 Abstract In the past years, the processing speed of the computer,storage power and network are all developing with an amazing speed . The great development provide a strong technical support for the digital image processing and network multimedia ,also expand the research and application.Image processing is becoming a booming science projects,it has a wide rang of reserching area and suject.Image processing theory and technology include:image digitization 、image enhancement 、image data compression 、image segmentation 、 feature extraction .The focus of this article is mage segmentation .First,image segmentation is the key steps from image processing to image analysis , and in the image works occupy an important place.But There are a lot Shortcomings than traditional methods,After this analysis we found the following findings:The traditional histogram threshold method The basic idea of the image segmentation is the image constitute the different feature regions, which show multi-peak histogram, each peak corresponding to a region to the threshold point valley by neighboring peaks.If the histogram valley obvious point, is relatively easy to select the threshold value, but in practice, due to the presence of noise, is often difficult to effectively select the threshold, the threshold value is set to sensitive noise. Wavelet-based image segmentationthreshold method is able to effectively avoid the noise. The basic idea is to firsttransform the image from the binary histogram is divided into different levels of wavelet coefficients, and then split according to a given wavelet coefficient selection criteria andthreshold threshold threshold marked the last use of the image segmentation area.Image segmentation with the previous side compared to greater advantage. Key words Wave；Haar Wave；Image Se