节点文献

复杂内河背景下天水线的提取

The sky-water line extraction under complicated inland waterway background

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 刘俊黄席樾何乃甩黄瀚敏

【Author】 LIU Jun,HUANG Xi-yue,HE Nai-shuai,HUANG Han-min(Laboratory of Navigation & Guidance,College of Automation,Chongqing University,Chongqing 400044,China)

【机构】 重庆大学自动化学院导航制导实验室

【摘要】 提出了一种在复杂内河背景下进行天水线提取的方法。该方法包括5个步骤:(1)利用均方误差(MSE)对原始图像进行评价;(2)根据步骤1的评价结果确定是否对图像进行预处理;(3)对经过步骤2处理的图像,分别利用算法1提取天水线得到天水线1,利用算法2提取天水线得到天水线2。算法1主要步骤有迭代阈值分割、Roberts梯度算子边缘检测、细化、Hough变换;算法2主要步骤有Roberts梯度算子边缘检测及二值化、细化、利用Hough变换;(4)把天水线1和天水线2中最靠近图像下端的那一条直线作为最终的天水线;(5)对最终提取的天水线进行主观定量评价,并使用证据理论进行可信度判断。实验结果表明,该方法适应性好,定位精度高,实时性和可靠性好。

【Abstract】 A method is presented to extract the sky-water line under complicated inland waterway background in this paper.The method consists five steps.Step 1:Mean squared error(MSE) is used to evaluate the original image;Step 2,the evaluation result from step 1 decides whether a pretreatment is need or not to the original image;Step 3,algorithm 1 and algorithm 2 are used to extract the sky-water line from the image which is treated by step 2 respectively.Sky-water line 1 is the result from algorithm 1 and sky-wwate line 2 is the result from algorithm 2.Algorithm 1 consists iterative threshold segmentation,edge detection using roberts gradient operators,thinning and hough transform.Algorithm 2 consists edge detection using roberts gradient operators and thresholding,Thinning and hough transform;Step 4,the sky-water line which is closer to the bottom of the image betwen sky-water line 1 and sky-water line 2 is regarded the final sky-water line;Step 5,The final sky-water line is assessed subjectively and quantificatinally,in addition,its credibility is judged by evidence theory.The result of the experiment show that the method has wide adaptability and high precision,and the method has fulfilled the demand of real-time and reliability.

  • 【文献出处】 光电子.激光 ,Journal of Optoelectronics.Laser , 编辑部邮箱 ,2008年08期
  • 【分类号】TP391.41
  • 【被引频次】7
  • 【下载频次】87
节点文献中: 

本文链接的文献网络图示:

本文的引文网络