LI Jing,XIAN Lin,WANG Haijiang.Research on Ship Detection Algorithm based on YOLOv3[J].Journal of Chengdu University of Information Technology,2023,38(01):37-43.[doi:10.16836/j.cnki.jcuit.2023.01.006]
基于YOLOv3的船只检测算法研究
- Title:
- Research on Ship Detection Algorithm based on YOLOv3
- 文章编号:
- 2096-1618(2023)01-0037-07
- Keywords:
- target detection; deep learning; a priori box; K-means
- 分类号:
- TP391.4
- 文献标志码:
- A
- 摘要:
- 针对长江流域错综复杂的生态环境以及执法部门人员短缺对长江10年禁渔令实施的限制情况; 通过智能视频监控系统对长江流域过往船只目标检测,对于判别船只有无违法捕捞行为具有重要意义。当前,传统的目标检测算法早已被检测效率更高、算法复杂度更低的深度学习方法替代; 基于智能视频监控对于实时性的要求,采用YOLOv3作为目标检测模型,在兼顾检测精度的同时检测速度也更高。YOLOv3算法中,先验框作为目标检测算法的重要机制,影响着预测框的定位性能。在K-means聚类算法上进行改进,通过改变K值初始化随机选择不能获取全局最优解的情况,对K值选择时应用轮盘法,选择距离已经形成的聚类中心尽可能远的值作为新的K值,使各个聚类中心相对距离尽可能大,从而尽可能获得全局最优的聚类结果。实验结果表明,K-means优化后获得的先验框训练模型让船只目标检测性能更加优异,整体mAP提升了9.31%。
- Abstract:
- In view of the intricate ecological environment of the Yangtze River Basin and the limitation of the shortage of law enforcement personnel on the implementation of the ten-year fishing ban in the Yangtze River; in this paper, the intelligent video surveillance system is used to detect the target of passing ships in the Yangtze River Basin, which is of great significance for judging whether there is illegal fishing behavior. At present, traditional target detection algorithms have long been replaced by deep learning methods with higher detection efficiency and lower algorithm complexity. This paper considers the real-time requirements of intelligent video surveillance, and adopts the YOLOv3 algorithm under comprehensive consideration, which takes into account the detection accuracy and the detection speed is also higher. In the YOLOv3 algorithm, the prior frame is an important mechanism of the target detection algorithm, which affects the positioning performance of the prediction frame. In this paper, the K-means clustering algorithm is improved. By changing the K value initialization, the random selection the global cannot obtain the optimal solution, the roulette method is applied to the selection of the K value, and the value that is as far as possible from the cluster center that has been formed is selected as the new K value, so that the relative distance of each cluster center is as large as possible, so as to obtain the globally optimal cluster as much as possible. The experimental results show that the K-means optimized prior frame training model makes the ship target detection performance more excellent, and the overall mAP is increased by 9.31%.
参考文献/References:
[1] 安建成.融合模糊LBP和Canny边缘的图像分割[J].计算机工程与设计,2019,40(12):3533-3537.
[2] 彭棉珠.卷积网络和SIFT特征融合的图像自动标注研究[J].福建电脑,2021,37(10):12-16.
[3] 贺瑜飞.基于Haar特征和改进的AdaBoost算法的人脸图像识别[J].榆林学院学报,2019,29(6):69-70.
[4] 兰胜坤.基于Adaboost算法的人脸检测实现[J].电脑与信息技术,2021,29(2):16-19.
[5] Felzenszwalb P F,Mcallester D A,Ramanan D.A discriminatively trained,multiscale,deformable part model[C].2008 IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2008.
[6] 胡葵,章东平,杨力.卷积神经网络的多尺度行人检测[J].中国计量大学学报,2017,28(4):472-477.
[7] Wen junyu,Sumi Kim,Jeong-Hyu Lee,et al.Animal Detection in Highly Cluttered Natural Scenes by using Faster R-CNN[J].International Journal of Recent Technology and Engineering(IJRTE),2019,8.
[8] Nirmala,S Arivalagan,R Arunkumar.An Efficient and Robust Multi-Object Recognition and Tracking Algorithm using Mask Region based Convolution Neural Network(R-CNN)[J].International Journal of Innovative Technology and Exploring Engineering(IJITEE),2019,8(9).
[9] K G Shreyas Dixit,Mahima Girish Chadaga, Sinchana S Savalgimath,et al.Evaluation and Evolution of Object Detection Techniques YOLO and R-CNN[J]. International Journal of Recent Technology and Engineering(IJRTE),2019,8.
[10] 龚强.基于Mask R-CNN的无人驾驶汽车道路前方目标检测的研究[D].南昌:南昌大学,2020.
[11] Ding H,Tian Y,Peng C,et al.Inference attacks on genomic privacy with an improved HMM and an RCNN model for unrelated individuals[J].Information Sciences,2020,512:207-218.
[12] He K,Zhang X,Ren S,et al.Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition[C].IEEE Transactions on Pattern Analysis & Machine Intelligence.2014:1904-1916.
[13] 焦李成.基于全局—局部SPP Net的高分辨SAR图像变化检测方法[D].西安:西安电子科技大学,2019.
[14] Zeghiche Oussama.基于深度Faster-RCNN的车牌识别算法研究[D].西安:西安电子科技大学,2020.
[15] 于晓倩.基于改进Faster RCNN的行人检测研究[D].长春:吉林大学,2020.
[16] Redmon J,Divvala S,Girshick R,et al.You Only Look Once:Unified,Real-Time Object Detection[C].Computer Vision & Pattern Recognition,2016.
[17] Shafiee M J,Chywl B,Li F,et al.Fast YOLO:A Fast You Only Look Once System for Real-time Embedded Object Detection in Video[J].Journal of Computational Vision and Imaging Systems,2017,3(1).
[18] 王婷.基于SSD的航拍图像中绝缘子识别与定位研究[D].北京:华北电力大学,2021.
[19] 姜敏,王力,王冬冬.改进的SSD行人检测算法[J].软件,2020,41(2):57-61.
相似文献/References:
[1]卢 丽,许源平,卢 军,等.基于社会力异常检测改进算法的人群行为模型[J].成都信息工程大学学报,2018,(01):1.[doi:10.16836/j.cnki.jcuit.2018.01.001]
LU Li,XU Yuan-ping,LU Jun,et al.A Crowd Behavior Model based on an ImprovedSocial Force Anomaly Detection Algorithm[J].Journal of Chengdu University of Information Technology,2018,(01):1.[doi:10.16836/j.cnki.jcuit.2018.01.001]
[2]胡 婕,陶宏才.基于深度学习的领域问答系统的设计与实现[J].成都信息工程大学学报,2019,(03):232.[doi:10.16836/j.cnki.jcuit.2019.03.004]
HU Jie,TAO Hongcai.Design and Implementation of Domain Question Answering System based on Deep Learning[J].Journal of Chengdu University of Information Technology,2019,(01):232.[doi:10.16836/j.cnki.jcuit.2019.03.004]
[3]王 强,李孝杰,陈 俊.基于He-Net的卷积神经网络算法的图像分类研究[J].成都信息工程大学学报,2017,(05):503.[doi:10.16836/j.cnki.jcuit.2017.05.007]
WANG Qing,LI Xiao-jie,CHEN Jun.Research on Image Classification based on HE-Net Convolutional Neural Networks[J].Journal of Chengdu University of Information Technology,2017,(01):503.[doi:10.16836/j.cnki.jcuit.2017.05.007]
[4]冉元波,孙 敏,高梦清,等.双偏振天气雷达水凝物识别研究[J].成都信息工程大学学报,2017,(06):590.[doi:10.16836/j.cnki.jcuit.2017.06.003]
RAN Yuan-bo,SUN Min,GAO Meng-qing,et al.Study on Hydrometeor Identification based on Deep Learning[J].Journal of Chengdu University of Information Technology,2017,(01):590.[doi:10.16836/j.cnki.jcuit.2017.06.003]
[5]周 咏,万 垚.基于无人机的监控系统设计[J].成都信息工程大学学报,2021,36(02):159.[doi:10.16836/j.cnki.jcuit.2021.02.006]
ZHOU Yong,WAN Yao.Design of Surveillance System based on UAV[J].Journal of Chengdu University of Information Technology,2021,36(01):159.[doi:10.16836/j.cnki.jcuit.2021.02.006]
[6]谭诗雨,杨 玲,师春香,等.复杂背景下银行卡号识别方法研究[J].成都信息工程大学学报,2021,36(03):280.[doi:10.16836/j.cnki.jcuit.2021.03.007]
TAN Shiyu,YANG Ling,SHI Chunxiang,et al.Bank Card Number Recognition System under the Complex Background based on Deep Learning[J].Journal of Chengdu University of Information Technology,2021,36(01):280.[doi:10.16836/j.cnki.jcuit.2021.03.007]
[7]郭楠馨,林宏刚,张运理,等.基于元学习的僵尸网络检测研究[J].成都信息工程大学学报,2022,37(06):615.[doi:10.16836/j.cnki.jcuit.2022.06.001]
GUO Nanxin,LIN Honggang,ZHANG Yunli,et al.Botnet Detection Method based on Meta-Learning Network[J].Journal of Chengdu University of Information Technology,2022,37(01):615.[doi:10.16836/j.cnki.jcuit.2022.06.001]
[8]魏春梅,马尚昌,卢会国,等.基于视频识别的气象观测场设备监控技术研究[J].成都信息工程大学学报,2023,38(02):129.[doi:10.16836/j.cnki.jcuit.2023.02.001]
WEI Chunmei,MA Shangchang,LU Huiguo,et al.Research on Equipment Monitoring Technology of Meteorological Observation Field based on Video Recognition[J].Journal of Chengdu University of Information Technology,2023,38(01):129.[doi:10.16836/j.cnki.jcuit.2023.02.001]
[9]毛 波,杨 昊,周世杰,等.基于CMA-REPS格点预报数据的深度学习风速订正方法[J].成都信息工程大学学报,2023,38(03):264.[doi:10.16836/j.cnki.jcuit.2023.03.003]
MAO Bo,YANG Hao,ZHOU Shijie,et al.A Deep Learning Method for Wind Speed Grid Point Forecasting Data Correction based on CMA-REPS[J].Journal of Chengdu University of Information Technology,2023,38(01):264.[doi:10.16836/j.cnki.jcuit.2023.03.003]
[10]任不凡,黄小燕,吴思东,等.基于语义信息的三维点云全景分割方法研究[J].成都信息工程大学学报,2023,38(05):535.[doi:10.16836/j.cnki.jcuit.2023.05.007]
REN Bufan,HUANG Xiaoyan,WU Sidong,et al.Research on Panoptic Segmentation of 3D Point Clouds based on Semantic Information[J].Journal of Chengdu University of Information Technology,2023,38(01):535.[doi:10.16836/j.cnki.jcuit.2023.05.007]
备注/Memo
收稿日期:2022-08-17