交通运输工程学报
交通運輸工程學報
교통운수공정학보
JOURNAL OF TRIFFIC AND TRANSPORTATION ENGINEERING
2015年
2期
118-126
,共9页
车辆检测%判别式模型%生成式模型%多模式弱分类器%AdaBoost-Bagging分类器
車輛檢測%判彆式模型%生成式模型%多模式弱分類器%AdaBoost-Bagging分類器
차량검측%판별식모형%생성식모형%다모식약분류기%AdaBoost-Bagging분류기
vehicle detection%discriminative model%generative model%multi-mode weak classifier%AdaBoost-Bagging classifier
针对现有车辆检测算法在实际复杂道路情况下对车辆有效检测率不高的问题,提出了融合多模式弱分类器,并以 AdaBoost-Bagging 集成为强分类器的车辆检测算法。结合判别式模型善于利用较多的特征形成较好决策边界和生成式模型善于利用较少的特征排除大量负样本的优点,以Haar 特征训练判别式弱分类器,以 HOG 特征训练生成式弱分类器,以 AdaBoost 算法为桥梁,采用泛化能力强的 Bagging 学习器集成算法得到 AdaBoost-Bagging 强分类器,利用 Caltech1999数据库和实际道路图像对检测算法进行了验证。验证结果表明:相比于单模式弱分类器,AdaBoost-Bagging 强分类器在分类能力和处理时间上均具有优越性,表现为较高的检测率与较低的误检率,分别为95.7%、0.00027%,每帧图像的检测时间较少,为25 ms;与传统级联 AdaBoost 分类器相比,AdaBoost-Bagging 强分类器虽然增加了12%的检测时间和30%的训练时间,但检测率提升了1.8%,误检率降低了0.00006%;本文算法的检测性能显著优于基于 Haar 特征的 AdaBoost 分类器算法、基于 HOG 特征的 SVM 分类器算法、基于 HOG 特征的 DPM 分类器算法,具有较佳的车辆检测效果。
針對現有車輛檢測算法在實際複雜道路情況下對車輛有效檢測率不高的問題,提齣瞭融閤多模式弱分類器,併以 AdaBoost-Bagging 集成為彊分類器的車輛檢測算法。結閤判彆式模型善于利用較多的特徵形成較好決策邊界和生成式模型善于利用較少的特徵排除大量負樣本的優點,以Haar 特徵訓練判彆式弱分類器,以 HOG 特徵訓練生成式弱分類器,以 AdaBoost 算法為橋樑,採用汎化能力彊的 Bagging 學習器集成算法得到 AdaBoost-Bagging 彊分類器,利用 Caltech1999數據庫和實際道路圖像對檢測算法進行瞭驗證。驗證結果錶明:相比于單模式弱分類器,AdaBoost-Bagging 彊分類器在分類能力和處理時間上均具有優越性,錶現為較高的檢測率與較低的誤檢率,分彆為95.7%、0.00027%,每幀圖像的檢測時間較少,為25 ms;與傳統級聯 AdaBoost 分類器相比,AdaBoost-Bagging 彊分類器雖然增加瞭12%的檢測時間和30%的訓練時間,但檢測率提升瞭1.8%,誤檢率降低瞭0.00006%;本文算法的檢測性能顯著優于基于 Haar 特徵的 AdaBoost 分類器算法、基于 HOG 特徵的 SVM 分類器算法、基于 HOG 特徵的 DPM 分類器算法,具有較佳的車輛檢測效果。
침대현유차량검측산법재실제복잡도로정황하대차량유효검측솔불고적문제,제출료융합다모식약분류기,병이 AdaBoost-Bagging 집성위강분류기적차량검측산법。결합판별식모형선우이용교다적특정형성교호결책변계화생성식모형선우이용교소적특정배제대량부양본적우점,이Haar 특정훈련판별식약분류기,이 HOG 특정훈련생성식약분류기,이 AdaBoost 산법위교량,채용범화능력강적 Bagging 학습기집성산법득도 AdaBoost-Bagging 강분류기,이용 Caltech1999수거고화실제도로도상대검측산법진행료험증。험증결과표명:상비우단모식약분류기,AdaBoost-Bagging 강분류기재분류능력화처리시간상균구유우월성,표현위교고적검측솔여교저적오검솔,분별위95.7%、0.00027%,매정도상적검측시간교소,위25 ms;여전통급련 AdaBoost 분류기상비,AdaBoost-Bagging 강분류기수연증가료12%적검측시간화30%적훈련시간,단검측솔제승료1.8%,오검솔강저료0.00006%;본문산법적검측성능현저우우기우 Haar 특정적 AdaBoost 분류기산법、기우 HOG 특정적 SVM 분류기산법、기우 HOG 특정적 DPM 분류기산법,구유교가적차량검측효과。
Focusing on the problem that the vehicle detection rate of existed vehicle detection algorithms is lower in real complex road environment, a vehicle detection algorithm was proposed,in which multi-model weak classifiers were integrated into strong classifier by using AdaBoost-Bagging method.In the algorithm,discriminative model could generate a fine decision boundary by using more features,and generative model could eliminate negative examples by using fewer features.To combine the advantages of discriminative model and generative model, discriminative classifier with Haar feature and generative classifier with HOG feature were built. Combined with AdaBoost algorithm,AdaBoost-Bagging strong classifier was obtained by using Bagging algorithm that is an integrated learning algorithm with strong generalization ability. Vehicle detection algorithm was tested based on Caltech1999 dataset and real road images.Test result indicates that compared with sole mode weak classifier,AdaBoost-Bagging strong classifier maintains superiority in classification ability and processing time,its high detection rate and low false detection rate are 95.7%,0.000 27% respectively,and the detection time of each frame is 25 ms that is less.Compared with the traditional cascade AdaBoost classifier,the detection time of the AdaBoost-Bagging strong classifier increases 12%,the training time increases 30%,but the detection rate increases 1.8%,and the false detection rate decreases 0.000 06%.The proposed algorithm is better than other vehicle detection algorithms,including Haar feature-based AdaBoost classifier, HOG feature-based SVM classifier, HOG feature-based DPM classifier,and has better vehicle detection effect.3 tabs,8 figs,25 refs.