摘要
图像分割是图像识别过程中的关键一步,分割结果的好坏会直接影响到星体的细分定位和模式识别。星体在星图中属于小目标,多目标,其分割与普通图像的分割有较大的不同。文中在分析星图特点的基础上,提出了基于星体面积的分割方法,并通过仿真与其他分割方法进行比较,验证了其有效性和优越性。
Image segmentation is a key step in image recognition. The effects of sub pixel location and star pattern recognition are directly affected by the result of segmentation. Star target in a star image is a small one. There are lots of stars in a star image. Thus, the segmentation of star image has many differences from an ordinary image. In this article, after analyzing the characteristic of a star image, a new method of star image segmentation based on the star＇s area was put forward and its validity and advantage were proved after emulation.
出处
《电子设计工程》
2012年第23期4-7,共4页
Electronic Design Engineering
基金
国家863计划创新基金(2010AAJ140)
关键词
星图分割
阈值分割
直方图
星体面积
star image segmentation
thresholding
histogram
area of star
