基于形态学运算的星图像分割

Star Image Segmentation Based on Morphology Algorithm

摘要:

在获取星图像的过程中,由于某些因素的影响,获得的星图像存在噪声和划痕,而且星图像的背景经常是不均匀的,为星图像的分割造成了极大的困难,甚至使得自动阈值分割无法实现或效果较差。膨胀和腐蚀是形态学的两个基本运算,实验证明,通过组合膨胀和腐蚀,可以获取更为复杂的形态学运算。用形态学运算对星图像进行处理,补偿不均匀的星图像背景,然后进行星图像的自动阈值分割可以得到很好的分割效果。

Abstract:

In the process of star image acquisition, there usually were kinds of noise and nicks in the acquired image because of some factors. In addition, the acquired star image often has an unequal background. Those characteristic makes the image segmentation difficult and even impossible, let alone automatic?image segmentation. Dilation and erosion were the two basic operations of morphology algorithm. More complicated operations can be got from different combination of dilation and erosion. Morphology algorithm can be used to compensate the background of star image. Segmentation based on the image whose background has been compensated has a much better effect.