基于HSV空间的视频实时水位检测算法  

An Algorithm of Real-time Water Level Detection via Video Based on HSV Space

作　　者：高晓亮[1] 王志良[1] 王馨[1] 刘冀伟[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083

出　　处：《郑州大学学报（理学版）》2010年第3期75-79,共5页Journal of Zhengzhou University:Natural Science Edition

基　　金：国家高技术发展计划(863)项目;编号2007AA10Z160

摘　　要：提出一种新的基于视频水位检测算法,能对设立于水中的标尺刻度进行自动和实时的测算.针对水面图像特有的属性,在将视频转换到HSV颜色空间的基础上,利用色调分量基本不受光照条件影响等特性,以一种基于能量函数的统计模型和可变区域的策略,解决了图像模糊、背景复杂且有倒影干扰等不利条件下,传统边缘检测算子无法准确测算的问题.A novel algorithm for water level detection is proposed based on HSV space via video.It has the ability to detect the scale on the ruler,which is set in the water automatically and in real-time.Aimed at the feature of water image,based on translating the video image into HSV space,and taking advantage of the feature that hue component is not affected by the condition of illumination on the whole,a strategy of variable domain approach and a model using energy function and statistical quantities are presented.Accordingly,the water level can still be detected accurately in the poor situation of blurring image and complicated background,which can not be solved by traditional edge detection operators.

关 键 词：水位检测 统计模型 颜色空间 能量函数 可变区域 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]