摘要
通过图像分割技术实时监测整个中药贴剂的生产流程,分析在线采集的贴剂图像的灰度特征、找到影响中药贴剂均匀度的因素,实现自动化生产。结合贴剂生产特点,采用基于灰度—梯度Otsu算法来提高对贴剂图像的分割精度;通过Otsu算法与粒子群优化算法结合,来降低图像分割的时间复杂度;针对影响因素的多样性,通过分离因子确定阈值分割个数。实验结果表明,该方法对贴剂图像具有良好的分割效果,满足速度和精度要求。
Image segmentation was used to monitor the production flow of Chinese patch in real-time.Gray feature provided the method to find the factor in order to achieve the automation production.With the feature of the patches production,this paper proposed an improved Otsu threshold selection method based on gray level-gradient two-dimensional histogram to improve the accuracy of segmentation.At the same time,used the particle swarm algorithm to search for the best threshold to reduce significant computation and meet the requirement of the real-time.According to the diversity of the affect factors,used separation factor to determine the number of the threshold segmentation.Experimental results demonstrate that the developed algorithm is fairly efficient,meets the requirements of speed and accuracy.
出处
《计算机应用研究》
CSCD
北大核心
2012年第1期359-362,共4页
Application Research of Computers
基金
国家科技重大专项课题(2009ZX09502)
关键词
图像分割
中药贴剂
均匀度
OTSU算法
粒子群优化算法
分离因子
image segmentation
Chinese medicine patch
uniformity
Otsu method
particle swarm optimization
separation factor