Fast binary image set operations on a run-based representation
Fast binary image set operations on a run-based representation
复制标题
基于运行的表示的快速二进制图像集操作
DOI：
10.1016/j.patrec.2016.06.029
复制
发表时间：
2016-09
影响因子：
5.1
通讯作者：
Liang Mengxia
中科院分区：
文献类型：
--
作者：
Guo Siyu;Zhou Weifang;Wen He;Liang Mengxia
集合运算是二值图像的常见处理。虽然通过简单的逐像素逻辑运算实现的集合运算通常是有效的，但存在所需集合运算的数量很大并且需要更快的集合运算的应用。针对这类应用，提出了一种基于游程的二值图像表示方法-游程森林，并在其上实现了常用的集合运算，如交、并、补、对称差和集合差等，游程森林是列的列表，也是图像列中游程的列表。两个运行的空间关系详尽列举。一个名为run iterator的数据结构被设计用来优雅地处理两次运行的操作。运行操作本身由逻辑操作和整数赋值组成，因此可以非常快。利用游程操作的简单性以及游程森林的可压缩性和有序性，有效地实现了集合操作。实验结果表明，虽然转换的二进制图像和从运行森林的计算开销，这可以在足够的集合运算的计算相当补偿，使所提出的方法是一个合适的选择，应用程序中的许多集合运算在很大程度上固定的二进制图像，或应用程序使用的运行森林作为基础表示整个。
Set operations are common processing of binary images. Though set operations implemented through naïve pixel-by-pixel logical operations are usually efficient, applications exist where the number of required set operations is large and faster set operations are needed. For such applications, a run-based representation, run forest, of binary images is proposed, and commonly used set operations of intersection, union, complementation, symmetric difference and set difference are realized on it. Run forests are lists of columns, which are also lists of runs in an image column. Spatial relations of two runs are exhaustively enumerated. A data structure called run iterator is designed to elegantly handle the operations of two runs. Run operations themselves consist of logical operations and assignments of integers, and can thus be very fast. Taking advantages of the simplicity of run operations, as well as the nature of run forest being compressive and well ordered, the set operations are efficiently realized. Experimental results show that although conversions of binary images to and from run forests cause computational overheads, this can be quite compensated during the computation of enough set operations, making the proposed method a suitable choice for applications with many set operations among largely fixed binary images, or applications using the run forest as the base representation throughout.
登录
查看更多内容
DOI：
10.1145/358826.358838
发表时间：
1980-03
期刊：
Commun. ACM
影响因子：
--
作者：
C. Dyer;A. Rosenfeld;H. Samet
通讯作者：
C. Dyer;A. Rosenfeld;H. Samet
一种基于信息融合的骨架剪枝算法
影响因子：
5.1
作者：
Liu Hongzhi;Wu Zhonghai;Zhang Xing;Hsu D. Frank
通讯作者：
Hsu D. Frank
DOI：
10.1016/0031-3203(94)00102-r
发表时间：
1995-03
期刊：
Pattern Recognit.
影响因子：
--
作者：
Chi-Yen Huang;K. Chung
通讯作者：
Chi-Yen Huang;K. Chung
DOI：
10.1016/s0167-8655(99)00007-0
发表时间：
1999-04
期刊：
Pattern Recognit. Lett.
影响因子：
--
作者：
D. Sarkar;N. Gupta
通讯作者：
D. Sarkar;N. Gupta
DOI：
10.1006/ciun.1994.1018
发表时间：
1994-05
期刊：
Cvgip: Image Understanding
影响因子：
--
作者：
W. D. Jonge;P. Scheuermann;A. Schijf
通讯作者：
W. D. Jonge;P. Scheuermann;A. Schijf