形态学变换的神经网及布尔函数实现

被引：1

作者：

陶鹏

李介谷

机构：

[1] 上海交通大学图像处理研究所

来源：

关键词：

形态支离; 前馈神经网表达; 布尔函数表达; 形态学变换;

D O I：

暂无

中图分类号：

TP183 [人工神经网络与计算];

学科分类号：

摘要：

本文提出了数学形态学算子的前馈神经网实现的模型.各种形态学算子(形态交离、腐蚀、膨胀、开启、闭合及多结构元素的基本形态运算的交或并)都可以根据算子本身置定权值、阈值和网结构以前馈神经网来实现.相应于对应算子和结构元素的神经网可通过简单的训练而直接得到.同时,我们也讨论了上述各种算子的布尔函数实现.

引用

页码：283 / 291

页数：9

共 11 条

[1] Greyscale morphology. Sternberg S R. CVGIP . 1986
[2] Image Analysis and Mathematical Morphology. Serra J. . 1982
[3] A note on the use of local min and max operations in digital picture processing. Y Nakagawa,A Rosenfeld. IEEE Transactions on Systems Man and Cybernetics . 1978
[4] The Algebraic Basis of Mathematical Morphology Part I: Dilations and Erosions. Heijmans H. J. A. M,,Ronse C. C.W.I, Amsterdam, The Netherlands. Rep. AM-R 8807 . 1988
[5] Beyond Mathematical Morphology. Ritter G X,Davidson J L,Wilson J N. Proc SPIE (Visual Comm Image Processing Ⅱ) . 1987
[6] Matheron,G. Random Sets and Integral Geometry . 1975
[7] E-filters. Preston K Jr. IEEE Transactions on Applied Superconductivity . 1983
[8] Neural Network Implementation of Mathematical Morphology Operation. Tao Peng,Li Jeigu. Proc SPIE (Applications of Artificial Neural Networks V) . 1994
[9] Image analysis using mathematical morphology. HARALICK R M,STERNBERG S R,ZHUANG X. IEEE Transactions on Pattern Analysis and Machine Intelligence . 1987
[10] A representation theory for morphological image and signal processing. Marogos P. IEEE Transactions on Pattern Analysis and Machine Intelligence . 1989