摘要
对活细胞进行精确计数,是体外抗病毒试验中测定药物抗病毒效果的重要手段。在进一步研究分析Gabor滤波器和所要分割的细胞图像特征的基础上,提出了带约束参数的Gabor滤波器,即根据Heisenberg测不准原理,用频域中高斯核的标准差来约束空域中高斯核的标准差;进而用标准差来约束Gabor滤波器中的中心频率的取值范围,达到提高效率和精度的目的。将带约束参数的Gabor滤波器用于细胞图像的增强,进而获得更高精度的分割和计数结果。对采集到的细胞图像进行大量实验,准确率达98.8%。
Accurate counting of living cells is an important measure of evaluating anti-virus effect of the drug in the ecto-antivirus experiment. Based on the study on Gabor filter and the features of cells to be segmented, parameter constrained Gabor filter was proposed in this paper. According to the Heisenberg uncertainty principle, the standard deviation of Gaussian kernel in the spatial domain was constrained by that in the frequent domain. Furthermore, the centre frequency of the Gabor filter is constrained by the standard deviation of the Gaussian kernel in the frequent domain to improve the efficiency and accuracy. A more accurate segmentation result was obtained by enhancing the cell images by parameter constrained Gabor filter. A large number of experiments gave the accurate ratio of 98.8 % .
出处
《中国生物医学工程学报》
CAS
CSCD
北大核心
2008年第3期335-339,377,共6页
Chinese Journal of Biomedical Engineering
基金
国家自然科学基金项目(60472047)
