一级黄色片免费播放|中国黄色视频播放片|日本三级a|可以直接考播黄片影视免费一级毛片

高級搜索

留言板

尊敬的讀者、作者、審稿人, 關(guān)于本刊的投稿、審稿、編輯和出版的任何問題, 您可以本頁添加留言。我們將盡快給您答復(fù)。謝謝您的支持!

姓名
郵箱
手機(jī)號碼
標(biāo)題
留言內(nèi)容
驗(yàn)證碼

基于深度置信網(wǎng)絡(luò)的隨機(jī)脈沖噪聲快速檢測算法

徐少平 張貴珍 李崇禧 劉婷云 唐祎玲

徐少平, 張貴珍, 李崇禧, 劉婷云, 唐祎玲. 基于深度置信網(wǎng)絡(luò)的隨機(jī)脈沖噪聲快速檢測算法[J]. 電子與信息學(xué)報(bào), 2019, 41(5): 1130-1136. doi: 10.11999/JEIT180558
引用本文: 徐少平, 張貴珍, 李崇禧, 劉婷云, 唐祎玲. 基于深度置信網(wǎng)絡(luò)的隨機(jī)脈沖噪聲快速檢測算法[J]. 電子與信息學(xué)報(bào), 2019, 41(5): 1130-1136. doi: 10.11999/JEIT180558
Shaoping XU, Guizhen ZHANG, Chongxi LI, Tingyun LIU, Yiling TANG. A Fast Random-valued Impulse Noise Detection Algorithm Based on Deep Belief Network[J]. Journal of Electronics & Information Technology, 2019, 41(5): 1130-1136. doi: 10.11999/JEIT180558
Citation: Shaoping XU, Guizhen ZHANG, Chongxi LI, Tingyun LIU, Yiling TANG. A Fast Random-valued Impulse Noise Detection Algorithm Based on Deep Belief Network[J]. Journal of Electronics & Information Technology, 2019, 41(5): 1130-1136. doi: 10.11999/JEIT180558

基于深度置信網(wǎng)絡(luò)的隨機(jī)脈沖噪聲快速檢測算法

doi: 10.11999/JEIT180558
基金項(xiàng)目: 國家自然科學(xué)基金(61662044, 61163023, 51765042, 81501560),江西省自然科學(xué)基金(20171BAB202017),江西省研究生創(chuàng)新項(xiàng)目(YC2018-S066)
詳細(xì)信息
    作者簡介:

    徐少平:男,1976年生,博士,教授,博士生導(dǎo)師,研究方向?yàn)閳D形圖像處理技術(shù)、機(jī)器視覺、虛擬手術(shù)仿真等

    張貴珍:女,1993年生,碩士生,研究方向?yàn)閳D像處理與機(jī)器學(xué)習(xí)

    李崇禧:男,1994年生,碩士生,研究方向?yàn)閳D像處理與機(jī)器學(xué)習(xí)

    劉婷云:女,1996年生,碩士生,研究方向?yàn)閳D像處理與機(jī)器學(xué)習(xí)

    唐祎玲:女,1977年生,博士生,研究方向?yàn)閳D像處理與機(jī)器學(xué)習(xí)

    通訊作者:

    唐祎玲 tangyiling@ncu.edu.cn

  • 中圖分類號: TP391

A Fast Random-valued Impulse Noise Detection Algorithm Based on Deep Belief Network

Funds: The National Natural Science Foundation of China (61662044, 61163023, 51765042, 81501560), The Project of Jiangxi Province Natural Science Foundation (20171BAB202017), The Jiangxi Provincial Graduate Innovation Special Fund (YC2018-S066)
  • 摘要:

    為提高現(xiàn)有隨機(jī)脈沖噪聲(RVIN)檢測算法的檢測準(zhǔn)確率和執(zhí)行效率,該文試圖從構(gòu)建描述能力更強(qiáng)的特征矢量和訓(xùn)練非線性映射更為準(zhǔn)確的預(yù)測模型兩個方面入手,實(shí)現(xiàn)一種基于訓(xùn)練策略的快速RVIN檢測算法。一方面,提取多個不同階的對數(shù)絕對差值排序統(tǒng)計(jì)值并結(jié)合一個能夠反映圖像邊緣特性的統(tǒng)計(jì)值作為刻畫圖塊中心像素點(diǎn)是否為噪聲的特征矢量。在計(jì)算量增加極少的情況下,顯著提升了特征矢量的描述能力。另一方面,基于深度置信網(wǎng)絡(luò)(DBN)訓(xùn)練RVIN預(yù)測模型(RVIN檢測器)將特征矢量映射為噪聲類型標(biāo)簽,實(shí)現(xiàn)了比淺層預(yù)測模型更為準(zhǔn)確的映射。大量實(shí)驗(yàn)數(shù)據(jù)表明:與現(xiàn)有的RVIN檢測算法相比,所提算法在檢測準(zhǔn)確率和執(zhí)行效率兩個方面都更有優(yōu)勢。

  • 圖  1  噪聲圖像中不同位置處2個圖塊中心像素點(diǎn)ROLD統(tǒng)計(jì)值比較

    圖  2  引入EF特征對噪聲檢測效果的影響比較

    表  1  圖1中a1和b1圖塊上所提取的前m階ROLD值比較

    圖塊階數(shù)m
    123456789101112
    a11.633.294.996.688.3710.0911.8013.5215.2416.9818.7220.46
    b11.002.203.665.126.848.6210.5012.4414.4016.4018.4220.45
    下載: 導(dǎo)出CSV

    表  2  各噪聲檢測算法在常用圖像集合上的各項(xiàng)性能指標(biāo)的平均值比較

    方法含噪20%含噪40%含噪60%
    漏檢數(shù)誤檢數(shù)錯檢總數(shù)MEMH漏檢數(shù)誤檢數(shù)錯檢總數(shù)MEMH漏檢數(shù)誤檢數(shù)錯檢總數(shù)MEMH
    ASWM3462106871414914.237478100051748316.401472098042452423.17
    PSMF1069535851427915.142303836032664130.273909656344473045.81
    ROLD-EPR656751061167318.77946289561841915.9510417116162203414.04
    ROR-NLM506893541442114.911190688732077917.1322553128563540823.60
    MLP-EPR850520811058622.781324457591900318.0915017101132513016.18
    本文方法40845909999211.77797585861656112.1710076125942267011.53
    下載: 導(dǎo)出CSV

    表  3  各檢測算法統(tǒng)一用相同修復(fù)算法降噪后在PSNR指標(biāo)上的比較(dB)

    方法含噪20%含噪40%含噪50%含噪60%
    LenaHouseBridgeLenaHouseBridgeLenaHouseBridgeLenaHouseBridge
    ASWM39.0633.3125.7634.2731.2124.3330.6628.8123.2526.0426.1321.61
    PSMF30.2427.8223.2529.2626.0322.7726.0324.0921.9122.0421.9820.00
    ROLD-EPR34.7733.3126.7531.7731.2124.2530.5428.8123.1228.7826.1322.20
    ROR-NLM36.9428.9225.2831.5828.9123.5927.6127.3922.3522.9224.6820.39
    MLP--EPR36.4539.3627.7133.8337.4924.4031.8636.4823.3329.4233.9622.25
    本文方法40.3542.9526.9735.8940.3124.9133.0438.2723.3629.6636.1222.18
    下載: 導(dǎo)出CSV

    表  4  各噪聲檢測算法平均執(zhí)行時(shí)間的比較(s)

    方法ASWMPSMFROLD-EPRROR-NLMMLP-EPR本文方法
    時(shí)間102.720.8610.4077.190.790.70
    下載: 導(dǎo)出CSV
  • LIU Licheng, Chen C L P, and ZHOU Yicong. A new weighted mean filter with a two-phase detector for removing impulse noise[J]. Information Sciences, 2015, 315(September): 4052–4057. doi: 10.1016/j.ins.2015.03.067
    SINGH N, THILAGAVATHY T, LAKSHMIPRIYA R T, et al. Some studies on detection and filtering algorithms for the removal of random valued impulse noise[J]. IET Image Processing, 2017, 11(11): 953–963. doi: 10.1049/iet-ipr.2017.0346
    DONG Yiqiu and XU Shufang. A new directional weighted median filter for removal of random-valued impulse noise[J]. IEEE Signal Processing Letters, 2007, 14(3): 193–196. doi: 10.1109/LSP.2006.884014
    劉萬軍, 梁雪劍, 曲海成. 自適應(yīng)增強(qiáng)卷積神經(jīng)網(wǎng)絡(luò)圖像識別[J]. 中國圖象圖形學(xué)報(bào), 2017, 22(12): 1723–1736. doi: 10.11834/jig.170079

    LIU Wanjun, LIANG Xuejian, and QU Haicheng. Adaptively enhanced convolutional neural network algorithm for image recognition[J]. Journal of Image and Graphics, 2017, 22(12): 1723–1736. doi: 10.11834/jig.170079
    于海平, 何發(fā)智, 潘一騰, 等. 一種基于多特征的距離正則化水平集快速分割方法[J]. 電子學(xué)報(bào), 2017, 45(3): 534–539. doi: 10.3969/j.issn.0372-2112.2017.003.004

    YU Haiping, HE Fazhi, PAN Yiteng, et al. A fast distance regularized level set method for segmentation based on multi-features[J]. Acta Electronica Sinica, 2017, 45(3): 534–539. doi: 10.3969/j.issn.0372-2112.2017.003.004
    王森, 王余, 王易川, 等. 水下高速目標(biāo)聲譜圖特征提取及分類設(shè)計(jì)[J]. 電子與信息學(xué)報(bào), 2017, 39(11): 2684–2689. doi: 10.11999/JEIT170283

    WANG Sen, WANG Yu, WANG Yichuan, et al. Feature extraction and classification of spectrum of radiated noise of underwater high speed vehicle[J]. Journal of Electronics &Information Technology, 2017, 39(11): 2684–2689. doi: 10.11999/JEIT170283
    周籮魚, 湯佳欣. 基于圖像稀疏表達(dá)的模擬退火圖像復(fù)原[J]. 光電子·激光, 2018, 29(2): 218–223. doi: 10.16136/j.joel.2018.02.0088

    ZHOU Luoyu and TANG Jiaxin. Image restoration based on simulated annealing with image sparse representation[J]. Journal of Optoelectronics · Laser, 2018, 29(2): 218–223. doi: 10.16136/j.joel.2018.02.0088
    AKKOUL S, LEDEE R, LECONGE R, et al. A new adaptive switching median filter[J]. IEEE Signal Processing Letters, 2010, 17(6): 587–590. doi: 10.1109/LSP.2010.2048646
    YU Hancheng, ZHAO Li, and WANG Haixian. An efficient procedure for removing random-valued impulse noise in images[J]. IEEE Signal Processing Letters, 2008, 15(12): 922–925. doi: 10.1109/LSP.2008.2005051
    GARNETT R, HUEGERICH T, CHUI C, et al. A universal noise removal algorithm with an impulse detector[J]. IEEE Transactions on Image Processing, 2005, 14(11): 1747–1754. doi: 10.1109/TIP.2005.857261
    DONG Yiqiu, RAYMOND H C, and XU Shufang. A detection statistic for random-valued impulse noise[J]. IEEE Transactions on Image Processing, 2007, 16(4): 1112–1120. doi: 10.1109/TIP.2006.891348
    XIONG Bo and YIN Zhouping. A universal denoising framework with a new impulse detector and nonlocal means[J]. IEEE Transactions on Image Processing, 2012, 21(4): 1663–1675. doi: 10.1109/TIP.2011.2172804
    ROY A, SINGHA J, DEVI S S, et al. Impulse noise removal using SVM classification based fuzzy filter from gray scale images[J]. Signal Processing, 2016, 128(11): 262–273. doi: 10.1016/j.sigpro.2016.04.007
    TURKMEN I. The ANN based detector to remove random-valued impulse noise in images[J]. Journal of Visual Communication and Image Representation, 2015, 34(10): 28–36. doi: 10.1016/j.jvcir.2015.10.011
    SOLEIMANY S and HAMGHALAM M. A novel random-valued impulse noise detector based on MLP neural network classifier[C]. Artificial Intelligence and Robotics (IRANOPEN), Qazvin, Iran, 2017: 165–169.
    HINTON G, DENG Li, YU Dong, et al. Deep neural networks for acoustic modeling in speech recognition: The shared views of four research groups[J]. IEEE Signal Processing Magazine, 2012, 29(6): 82–97. doi: 10.1109/MSP.2012.2205597
    WANG Zhou and ZHANG D. Progressive switching median filter for the removal of impulse noise highly corrupted images[J]. IEEE Transactions on Circuits System II: Analog Digital Signal Processing, 1999, 46(1): 78–80. doi: 10.1109/82.749102
    CHAN S H, WANG Xiran, and ELGENDY O A. Plug-and-play ADMM for image restoration: Fixed-point convergence and applications[J]. IEEE Transactions on Computational Imaging, 2017, 3(1): 84–98. doi: 10.1109/TCI.2016.2629286
  • 加載中
圖(2) / 表(4)
計(jì)量
  • 文章訪問數(shù):  2932
  • HTML全文瀏覽量:  934
  • PDF下載量:  92
  • 被引次數(shù): 0
出版歷程
  • 收稿日期:  2018-06-06
  • 修回日期:  2018-12-07
  • 網(wǎng)絡(luò)出版日期:  2018-12-13
  • 刊出日期:  2019-05-01

目錄

    /

    返回文章
    返回