摘要
运动异物的正确分割是实现液剂自动检测的关键.提出一种基于帧差图空间方差和方差梯度的液剂异物模糊自适应阈值分割算法.首先将4帧序列图像差分得到的2幅帧差图分为5×5的图像块,计算2幅帧差图对应图像块间的方差和方差梯度,并以两者的乘积构成新特征图;然后使用自适应阈值对新特征图进行运动异物分割和提取.为使阈值能够跟随图像块灰度变化,阈值的调整采用模糊推理依据方差和方差梯度变化自适应实现.实验及实际测试结果表明,该算法能够满足低对比度和局部光照变化的液剂异物实时检测要求,是一种实用有效的图像分割算法.
The effective segmentation of moving impurity in liquid plays a key role in automatic detecting system. A fuzzy adaptive segmentation algorithm of moving impurity in liquid based on spatial block variance and its gradient is proposed. At first, two difference images from 4 consecutive frames are partitioned into 5 × 5 small blocks respectively, and then the homologous block variance and its gradient are computed, and their product is used as the block feature to construct a new image. After that, an adaptive thresholding is used to segment the new image. To cope with the changes of the block grey value, a fuzzy inference method is adopted to adjust the threshold adaptively according to the block variance and its gradient. The experimental and factual testing results show that the proposed algorithm is practical, effective, and can meet the demands of real-time detection of moving impurity in liquid
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2009年第8期1144-1148,1154,共6页
Journal of Computer-Aided Design & Computer Graphics
基金
国家自然科学基金(50575168)
陕西省教育厅自然科学专项基金(07JK281)
关键词
液剂异物
图像分割
方差
方差梯度
模糊自适应阈值
threshold impurity in liquid
image segmentation
variance
variance gradient
fuzzy adaptive
