MATLAB杂草图像处理技术附研究报告论文.docxVIP

本科毕业论文 杂草图像处理技术及应用 学 院： 计算机与信息工程学院 专 业： 计算机科学与技术专业 学 号： 071010376 姓 名： 乌文泽 指导教师： 薛河儒 职 称： 教授 论文提交日期：二О一一年六月 摘 要 图像分割是图像处理中的一项关键技术，其目的是将目标和背景分离，为后续处理提供依据，其结果直接影响到其后的信息处理过程。到目前为止，对图像分割的效果好坏，还没有一个统一的评价判断准则。不同的分割方法对同一幅图像的分割效果是不同的。本文研究不同分割方法对同一副图像进行分割后的定量比较。 灰度阈值法是一种常使用的分割方法。图像阈值化处理的实质是一种图像灰度级的非线性运算，阈值处理可用方程加以描述。通过选取阈值将目标与背景分离出来。 分水岭变换是一种数学形态学图像分割方法，它可以得到精确的边缘，即连续、封闭、单像素宽的边缘。但是它的主要缺点就是对噪声十分敏感，很弱的噪声就会造成严重的过分割现象，致使分割结果出现大量的零散区域。 本文是用实验的方法，针对农田植物与土壤对比比较强烈的情况，将其背景与植物分离。利用不同算法对同一幅图像进行分割，然后对其效果进行定量比较，从而确定一种分割效果较好的方法。 关键词：图像分割 颜色特征 灰度阈值法 分水岭算法 定量评价 Abstract Image segmentation is a key technology of image processing, and its purpose is to separate the target from the background. It makes further image analysis and comprehension possible. But so far, there is not a unified evaluation criterion about the effects of image segmentation. The effects of segmentation are different with different segmentation methods, and the effects are also different with a given segmentation method at different color spaces. This paper studies different segmentation method for the same vice image segmentation quantitative comparison after. Gray threshold value method is a frequently used segmentation method. Image threshold processs essence is a kind of image grayscale nonlinear operations, threshold processing can be used to describe equation. By choosing threshold will target and background is isolated The watershed is an image segmentation method based on mathematical morphology. It gets the precise edge which is continuous, closed and single-pixel. The main limitation of watershed transform is the over-segmentation due to its sensitivity to noise; even the very thin noise will lead to a lot of scattered and meaningless regions. This paper is to use the method of experimental comparison for farmland plants and soil more intense, will its background and plant separation. Use different algorithms on the same image segmentation, then its effect, thus determine quantitatively c