• |本期目錄/Table of Contents|

    [1]張秀玲,魏其珺,周凱旋,等.基于改進LeNet-5網絡的車牌字符識別[J].沈陽大學學報,2020,(4):312-317.
    點擊復制

    基于改進LeNet-5網絡的車牌字符識別

    《沈陽大學學報》[ISSN:1000-9035/CN:22-1262/O4]

    卷:
    期數:
    2020年第4期
    頁碼:
    312-317
    欄目:
    信息與控制
    出版日期:
    2020-08-25

    文章信息/Info

    Title:
    -
    作者:
    張秀玲魏其珺周凱旋董逍鵬馬 鍇
    燕山大學 河北省工業計算機控制工程重點實驗室,河北 秦皇島 066004
    Author(s):
    -
    關鍵詞:
    卷積神經網絡車牌字符識別LeNet-5網絡Inception-SE卷積模塊識別精度
    分類號:
    -
    DOI:
    -
    文獻標識碼:
    -
    摘要:
    引入了Inception-SE卷積模塊組來提升 LeNet-5網絡的廣度與深度,運用 SE模塊增強了有用 的特征并抑制了對當前任務用處不大的特征;使用 BN 層和 Dropout優化網絡,防止梯度彌散,提升精度;使 用全局池化層(globalaveragepooling,GAP)代替全連接層來減少網絡計算參數.研究結果表明:改進后網絡 的識別精度達到了99.88%,比傳統的 LeNet-5網絡提高了1.71%.

    參考文獻/References

    [1] RASHIDAE.Afastalgorithmforlicenseplatedetection [C]∥2013InternationalConferenceonSignalProcessing, ImageProcessing & PatternRecognition,February7-8, 2013.Coimbatore.IEEE,2013:44 48.
    [2]張秀玲,齊晴,侯代標,等.混合優化的 RBF網絡車牌字符 識別[J].沈陽大學學報(自然科學版),2017,29(2):113 117. ZHANG X L,QI Q,HOU D B,etal.Licenseplate recognition of RBF neutral network based on mixed optimization[J].JournalofShenyangUniversity(Natural Science),2017,29(2):113 117.
    [3] YAOD H,ZHU W X,CHEN YJ,etal.Chineselicense platecharacterrecognition based onconvolution neural network [C]∥ 2017 Chinese Automation Congress (CAC),October20-22,2017.Jinan.IEEE,2017:1547 1548.
    [4] CHEN H L,HU B,YANG X,etal.Chinesecharacter recognitionfor LPR application[J].Optik,2014,125 (18):5295 5302.
    [5] ZHU Y Q,LIC H.Arecognition methodofcarlicense plate characters based on template matching using modified Hausdorff distance[C]∥ 2010 International Conference on Computer, Mechatronics, Control and ElectronicEngineering,August24-26,2010.Changchun, China.IEEE,2010:25 28.
    [6] YANGDN,KONGJ,DU N,etal.Anovelapproachfor licenseplatelocalizationbasedonSVM classifier[C]∥ 20102ndIEEEInternationalConferenceonInformation Management and Engineering, April 16-18, 2010. Chengdu,China.IEEE,2010:1 6.
    [7] 張培玲,畢東生,資麗.基于改進 BP神經網絡的夜間車牌 識別算法研究[J].測控技術,2018,37(8):21 24. ZHANGP L,BID S,ZIL.Researchonnightvehicle licenseplaterecognitionalgorithm basedonimprovedBP neuralnetwork[J].Measurement& ControlTechnology, 2018,37(8):21 24.
    [8] WUJ,XIAOZT.Newlicenseplatecharacterrecognition algorithm based on ICM [C]∥ 2011 International Conference on Control, Automation and Systems 316 沈陽大學學報(自然科學版) 第32卷 Engineering,July30-31,2011,Singapore.IEEE,2011:1 4.
    [9] LIU Y J,HUANG H,CAO J D,etal.Convolutional neuralnetworks-basedintelligentrecognitionofChinese licenseplates[J].SoftComputing,2018,22(7):2403 2419.
    [10] 王海洲,王冠.基于深度學習的人工智能醫療應用與存在 的問題[J].沈陽大學學報(自然科學版),2019,31(2): 118 121. WANG H Z,WANG G. Application and existing problemsofartificialintelligence medicalbasedondeep learning[J].JournalofShenyang University(Natural Science),2019,31(2):118 121.
    [11] 趙春暉,周瑤.基于改進 FasterR-CNN 算法的艦船目標 檢測與識別[J].沈陽大學學報(自然科學版),2018,30 (5):366 371. ZHAO C H,ZHOU Y. Ship target detection and recognitionbased onimprovedfaster RCNN algorithm [J].JournalofShenyang University(NaturalScience), 2018,30(5):366 371.
    [12] YIN J W,LYU X Y,WEI T,etal.Face-password certificationbasedonconvolutionalneuralnetwork:AU 2019100349[P].2019 05 04.
    [13] ABDEL-HAMID O,MOHAMED A R,JIANG H,etal. Convolutionalneuralnetworksforspeechrecognition[J]. ACM Transactions on Audio,Speech,and Language Processing,2014,22(10):1533 1545.
    [14] TAN S,TAN Z.Improved LeNet-5 modelbased on handwritten numeralrecognition [C]∥ 2019 Chinese Controland Decision Conference (CCDC),June 3-5, 2019,Nanchang.IEEE,2019:6396 6399.

    備注/Memo

    備注/Memo:
    收稿日期:2019 12 31
    基金項目:國家自然科學基金資助項目(61573303).
    作者簡介:張秀玲(1968 ),女,山東章丘人,教授,博士.
    更新日期/Last Update: 2020-08-22
    开心赛车