NSCT域內(nèi)結(jié)合邊緣特征和自適應(yīng)PCNN的紅外與可見光圖像融合
發(fā)布時間:2018-08-14 15:37
【摘要】:針對傳統(tǒng)的基于多尺度變換的紅外與可見光圖像融合,對比度不高,邊緣等細節(jié)信息保留不充分等問題,結(jié)合NSCT變換的多分辨率、多方向特性和PCNN全局耦合、脈沖同步激發(fā)等優(yōu)點,提出一種基于NSCT變換結(jié)合邊緣特征和自適應(yīng)PCNN紅外與可見光圖像融合算法.對于低頻子帶,采用一種基于邊緣的融合方法;對于高頻方向子帶,采用方向信息自適應(yīng)調(diào)節(jié)PCNN的鏈接強度,使用改進的空間頻率特征作為PCNN的外部激勵,根據(jù)脈沖點火幅度融合子帶系數(shù).實驗結(jié)果驗證了該算法的有效性.
[Abstract]:Aiming at the problems of traditional infrared and visible image fusion based on multi-scale transform, the contrast is not high, the detail information such as edge is not enough, and so on, the multi-resolution, multi-direction characteristic and PCNN global coupling of NSCT transform are combined. Based on NSCT transform and edge feature, an adaptive PCNN infrared and visible image fusion algorithm is proposed. For low-frequency subbands, an edge-based fusion method is used, and for high-frequency directional subbands, directional information is used to adaptively adjust the link strength of PCNN, and an improved spatial frequency feature is used as the external excitation of PCNN. According to the pulse ignition amplitude fusion subband coefficient. Experimental results show that the algorithm is effective.
【作者單位】: 武漢大學(xué)測繪學(xué)院;武漢大學(xué)測繪遙感信息工程國家重點實驗室;
【基金】:國家自然科學(xué)基金(No.41271456)
【分類號】:TP391.41
[Abstract]:Aiming at the problems of traditional infrared and visible image fusion based on multi-scale transform, the contrast is not high, the detail information such as edge is not enough, and so on, the multi-resolution, multi-direction characteristic and PCNN global coupling of NSCT transform are combined. Based on NSCT transform and edge feature, an adaptive PCNN infrared and visible image fusion algorithm is proposed. For low-frequency subbands, an edge-based fusion method is used, and for high-frequency directional subbands, directional information is used to adaptively adjust the link strength of PCNN, and an improved spatial frequency feature is used as the external excitation of PCNN. According to the pulse ignition amplitude fusion subband coefficient. Experimental results show that the algorithm is effective.
【作者單位】: 武漢大學(xué)測繪學(xué)院;武漢大學(xué)測繪遙感信息工程國家重點實驗室;
【基金】:國家自然科學(xué)基金(No.41271456)
【分類號】:TP391.41
【相似文獻】
相關(guān)期刊論文 前10條
1 鄒文潔;;基于PCNN神經(jīng)網(wǎng)絡(luò)的圖像去噪算法研究[J];計算機仿真;2008年08期
2 齊永鋒;火元蓮;張家樹;;基于簡化的PCNN與類內(nèi)最小離散度的圖像自動分割方法[J];光電子.激光;2008年09期
3 王念一;馬義德;綻琨;;PCNN理論研究進展及其語音識別中的應(yīng)用[J];自動化與儀器儀表;2013年01期
4 鐘亞平;;基于PCNN的手腕骨圖像自動分割方法研究[J];軟件導(dǎo)刊;2013年07期
5 顧曉東,郭仕德,余道衡;一種基于PCNN的圖像去噪新方法[J];電子與信息學(xué)報;2002年10期
6 趙峙江;趙春暉;張志宏;;一種新的PCNN模型參數(shù)估算方法[J];電子學(xué)報;2007年05期
7 火元蓮;;一種基于PCNN的圖像自動分割方法[J];自動化與儀器儀表;2008年06期
8 李敏;蔡騁;談?wù)?;基于修正PCNN的多傳感器圖像融合方法[J];中國圖象圖形學(xué)報;2008年02期
9 劉R,
本文編號:2183333
本文鏈接:http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2183333.html
最近更新
教材專著