液体泡沫中单气泡的图像识别算法  

Single bubbles recognition algorithm in liquid foam image

作　　者：康乐[1] 冯殿义[1] 胡瑢华[2] 

机构地区：[1]辽宁工业大学机械工程与自动化学院,辽宁锦州121001 [2]南昌大学机械电子工程学院,江西南昌330031

出　　处：《渤海大学学报（自然科学版）》2017年第2期167-172,共6页Journal of Bohai University:Natural Science Edition

基　　金：辽宁工业大学高校青年教师启动基金项目(No:X201104)

摘　　要：根据泡沫中气泡的特征,提出了一种液体中单气泡的图像识别算法.在垂直光源拍摄位置获取清晰的泡沫图像.对原始泡沫图像进行全局阈值处理和形态学处理过,获得处理后的二值化图像.为了防止轮廓提取产生死循环,提出了一种毛刺去除方法.轮廓提取后,利用Hough变换对每个单连通边缘识别是否为类圆形,对圆心和半径进行投票,对投票结果进行分析,用高斯分布函数来估计投票峰值点.利用本文方法识别到泡沫图像中的所有单气泡.最后,对实现结果进行分析,证明了这种单气泡识别方法效果较理想.According to the characteristics of the foam bubbles, a new single bubble identification algorithm in liquid image is proposed. This paper gets a clear picture of the foam on the vertical light shooting position. The original foam image is processed by globally thresholding and morphological processing, and then a binary image is obtained. In order to prevent generation loop in contour extraction, this paper proposes a method of removing burrs. After contour extraction, Hough transform is used to identify whether each single communication edge is round like, which by voting for the center and radius, analyzing the voting results and estimating the peak point of the vote from Gaussian distribution function. The method in this paper can be used to identify all single bubbles in the foam image. Finally, analyzing the achieved results, it shows that this single bubble identification method is effect ideal.

关 键 词：全局阈值处理 形态学图像处理 HOUGH变换 高斯分布 

分 类 号：TF068[冶金工程—冶金物理化学]