圖像濾波的形態(tài)學(xué)開(kāi)、閉型神經(jīng)網(wǎng)絡(luò)算法
MORPHOLOGICAL OPENING AND CLOSING NEURAL NETWORKS FOR IMAGE FILTERING
-
摘要: 該文設(shè)計(jì)完成了一種具有實(shí)用意義的形態(tài)學(xué)開(kāi)、閉濾波的神經(jīng)網(wǎng)絡(luò)模型及其濾波參數(shù)的優(yōu)化訓(xùn)練算法。實(shí)驗(yàn)結(jié)果表明該方法設(shè)計(jì)簡(jiǎn)便,實(shí)用性強(qiáng)且易于推廣,對(duì)提高形態(tài)濾波性能效果明顯。分析表明,形態(tài)濾波器可分解為形態(tài)濾波運(yùn)算和結(jié)構(gòu)元素選擇兩個(gè)基本問(wèn)題。形態(tài)濾波運(yùn)算規(guī)則已由定義本身確定,于是形態(tài)濾波器的最終濾波性能就僅僅取決于結(jié)構(gòu)元素的選擇。進(jìn)行自適應(yīng)優(yōu)化訓(xùn)練的目的正是使結(jié)構(gòu)元素具有圖像目標(biāo)的形態(tài)結(jié)構(gòu)特征,從而使形態(tài)濾波器對(duì)復(fù)雜變化的圖像具有良好的濾波性能和穩(wěn)健的適應(yīng)能力。Abstract: This paper presents morphological neural networks of opening and closing operation for pratical use, and the algorithm to design optimal parameters of a morphological filter, Experimental results show that this method is good in practice and easy to extend. It has better filtering properties than that of the conventional morphological ones. The task of creating a morphological filter can be divided into two basic problems, selecting a morphological operation and Structuring Element (SE). The set of morpholo...
-
E.R. Dougherty, et al., Digital Image Processing Methods, New York, Marcel Dekker, 1994,110-138.[2]龔煒,石青云,程民德,數(shù)字空間中的數(shù)學(xué)形態(tài)學(xué),北京,科學(xué)出版社,1997,137-162.[3]C.P. Suarez-Araujo, Novel neural-network models for computing homothetic invariances: Animage algebra notation, J. Math. Imaging and Vision, 1997, 7(1), 69-83.[4]Yonggwan Won, et al., Morphological shared-weight networks with applications to automatictarget recognition, IEEE Trans. on Neural Networks, 1997, NN-8(5), 1195-1203.[5]Gerhard X. Ritter, Morphological associatiative memories, IEEE Trans. on Neural Networks,1998, NN-9(2), 281-292.[6]余農(nóng),李吉成,一種自適應(yīng)訓(xùn)練形態(tài)濾波參數(shù)的神經(jīng)網(wǎng)絡(luò)方法,中國(guó)電子學(xué)會(huì)第四屆青年學(xué)術(shù)年會(huì)文集,北京,電子工業(yè)出版社,1998,537-540.[7]李吉成,李飚,灰度形態(tài)濾波器的神經(jīng)網(wǎng)絡(luò)實(shí)現(xiàn)方法,系統(tǒng)工程與電子技術(shù),1999,21(3),56-59.[8]汪云九,崔,齊翔林,BP學(xué)習(xí)網(wǎng)絡(luò)中權(quán)值的感受野型初始化研究,自然科學(xué)進(jìn)展,1996,6(3),346-350. -
計(jì)量
- 文章訪問(wèn)數(shù): 2358
- HTML全文瀏覽量: 123
- PDF下載量: 520
- 被引次數(shù): 0
下載:
