计算机应用 ›› 2010, Vol. 30 ›› Issue (12): 3274-3277.

• 图形图像处理 • 上一篇    下一篇

内皮细胞图像自动分割与荧光强度测量方法

郭斯羽1,鲍美华2,翟文娟1,唐求1   

  1. 1. 湖南大学电气与信息工程学院
    2. 长沙医学院
  • 收稿日期:2010-05-19 修回日期:2010-07-08 发布日期:2010-12-22 出版日期:2010-12-01
  • 通讯作者: 郭斯羽
  • 基金资助:
    湖南大学青年教师科技创新扶持项目

Automatic segmentation and fluorescent intensity measurement for endothelial cell images

  • Received:2010-05-19 Revised:2010-07-08 Online:2010-12-22 Published:2010-12-01

摘要: 针对内皮细胞提出了一种自动分割与细胞区域荧光强度测量方法。通过形态学重建和背景减除消除图像的光照不均匀背景;利用直方图均衡化进行图像增强;增强图像经自动阈值分割与形态学滤波得到二值化粗分割结果;在粗分割结果上利用形态学细化和分水岭算法获取前景与背景的标记点;结合增强图像的梯度及标记点,利用标记点控制的分水岭算法完成分割;最终通过求取分割所得细胞区域内的平均荧光强度完成测量。在实际图像上的实验结果表明,相比于直接使用阈值分割方法,所提出的方法能更准确地完成图像分割及荧光强度的测量。

关键词: 形态学重建, 形态学滤波, 阈值分割, 分水岭算法, 荧光强度测量

Abstract: Quantitative measurement of fluorescent emission is important. An automatic segmentation and cell region fluorescent intensity measuring method was proposed for endothelial cells. Morphological image reconstruction and background subtraction were used to eliminate the uneven image background. Histogram equalization was adopted for image enhancement. Automatic thresholding and morphological filtering were subsequently carried out on the enhanced image, resulting in the binary coarse segmentation, which was afterwards processed by morphological thinning and watershed transform to obtain foreground and background markers, respectively. Marker-controlled watershed was performed on the gradient map of the enhanced image to give the final segmentation. Average fluorescent intensity over the segmented cell regions was calculated and the measurement was completed. The experimental results on real-world images show that, compared with the direct thresholding approach, the proposed method can achieve more accurate segmentation and fluorescent intensity measurement.

Key words: morphological reconstruction, morphological filtering, threshold segmentation, watershed algorithm, fluorescent intensity measurement