基于阈值的彩色目标图像的提取与测量

幅彩色图像提供了很多不同的信息，但是为了更好地研究我们所感兴趣的目标图像，需要将它们以图像斑块的形式单独提取出来，从而排除研究所不需要的干扰信息和背景信息，使得研究上作更容易、更准确。通过合理的选取阈值及提取关系，如大于等于阈值、小于等于阈值等，可从图像中提取研究所关心的目标图像及其边界，并测量所提取的目标图像的总面积及目标图像各斑块的面积和周长，由此可研究其分布规律与分布密度；并可在测量的基础上对目标图像斑块进行统计，统计出在一定面积或周长范围内目标图像斑块的个数，为相应的研究提供定量分析。

Extraction and measure of the colored target image based on threshold value

A colored image may provide a lot of different information, but in order to researching the interested target image better, we need to extract the target image solely by the form of image speckles, thus remove the needless intefferential information and the background information to make the research is more accurate and easy. We can extract target image and iti＇-s border that the research concern from the image through choosing threshold value and extraction relation, such as not less than threshold value, not more than threshold value etc, and measure total area of the target image and area and perimeter of each image speckles of the target image. So we can find the distributing disciplinarian and density of the target image from these. We can compute the target image speckles numbers within some scopes of area and perimeter to provide quantitative analysis for the homologous research.