基于水平集的绿化区域分割方法研究计算机应用技术专业论文.docxVIP

I II Abstract Image segmentation is the key link of the field of image processing, has important applications in medical image processing, video image processing, remote sensing, satellite image processing. On the basis of remote sensing image preprocessing, image segmentation method can effectively extract images of green vegetation, you can explore the application prospects of Chinas ecological environment construction, so as to promote Chinas western vegetation restoration and ecological environment construction, monitoring and assessment of desertification status quo Forestry key construction and sand control effect, forest and grassland fires. In addition, clear and intuitive observed the status quo of the greening of the city, to facilitate the development of co-ordination and reasonable greening policy, establish a green ring urban ecological environment, strengthen the greening of the residential area, and improve the quality of life of the people. Classical active contour models based on the area is the Chan-Vese model that overcomes the image gradient information segmentation results with certain anti-noise ability. Traditional C-V model, the use of the global scope of the image, with the global dependence segmentation performance will affect the range of options and the image information. This paper studies the relationship between the image the high-frequency component of the Information and split dependent range of information and C-V level set based on wavelet local energy remote sensing image segmentation algorithm is presented. The method of traditional C-V level set model into a global information-dependent information based on wavelet local energy-dependent segmentation method to improve the accuracy of the complex background of remote sensing image segmentation. In addition, C-V model is a scalar field model, using only the image intensity and texture image segmentation is a problem in image processing, the precise target contour for the complex