基于连通区域标记的区域填充算法  被引量:8

Regional Filling Algorithm Based on Connected Region Labeling

在线阅读下载全文

作  者:苗龙元 于正林[1] 王震[1] MIAO Longyuan;YU Zhenglin;WANG Zhen(School of Mechanical Engineering,Changchun University of Science and Technology,Changchun 130022)

机构地区:[1]长春理工大学机电工程学院,长春130022

出  处:《长春理工大学学报(自然科学版)》2018年第4期114-117,共4页Journal of Changchun University of Science and Technology(Natural Science Edition)

摘  要:针对图像中外形复杂的区域难以填充问题,提出一种基于连通区域标记的区域填充算法对复杂区域进行填充。该算法首先利用连通区域标记实现对二值图像的分区标记,其次对标记矩阵L最大行(列)、最小行(列)上的元素进行检测,最终通过对检测到的区域结果对错误填充的部分求反,实现区域填充。相比扫描线填充算法,该算法无需对边界进行跟踪和复杂分类,填充效果更为理想。与种子算法相比,该算法不存在重复填充,无需进行颜色标记,同时处理速度更快。通过MATLAB实验证明,基于连通区域标记的区域填充算法可以填充外形任意复杂的区域,且填充效率高,精度高等特点,适用于超大分辨率图像的填充,为填充复杂区域提供了一种新思路。Aiming at the problem that the area with complex shape was difficult to fill in the image,a region filling algorithm based on connected region labeling was proposed to fill the complex area. Firstly,the partition of binary image was labeled by using connected region labeling algorithm to. Secondly,all elements of the maximum and minimum row(column) of the tagged matrix L were detected. Finally,based on the detected region results to reverse the wrong areas regional filling was realized. Compared with the scanning line filling algorithm,the algorithm does not need to track and classify the boundary,and the filling effect is more ideal. Compared with seed algorithm,the algorithm does not have repeated filling,no color coding is required,and the processing speed is faster. Through MATLAB experiments show that regional filling algorithm based on connected component labeling can fill any complex shape of the area with high filling efficiency and high precision filling characteristics,is suitable for filling of super resolution image to fill the complex area and provides a new way of filling Region.

关 键 词:区域填充 区域连通标记 任意区域 效率高 超大分辨率图像 

分 类 号:TP391.41[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象