一种实时的道路空车位检测算法
A Real-Time Road Empty Parking Space Detection Algorithm
摘要: 道路空车位实时检测是解决城市停车难的一个关键环节，本文提出了一种改进的faster RCNN深度学习道路空车位检测方法。首先，利用K-means聚类算法对标注框进行聚类，以获得适合的空车位形状特征；然后，调整区域建议框网络(Regional proposals network)框架，以便网络提取更多的数据特征；最后，利用soft NMS算法提取可能包含目标的框。实验结果表明，建议的方法对道路空车位的检测结果较好,在采集的道路空车位检测数据集准确率达到95.3%，比只用faster RCNN方法提高了2个百分点。
Abstract: The real-time detection of road empty parking spaces is a key link to solve the difficulty of urban parking. This paper proposes an improved faster RCNN deep learning road empty parking space detection method. First, use the K-means clustering algorithm to cluster the labeled boxes to obtain suitable vacant parking space shape features; then, adjust the regional proposal box network (RPN) framework so that the network can extract more data features; finally, use soft The NMS algorithm extracts boxes that may contain targets. The experimental results show that the proposed method has better detection results for road empty parking spaces. The accuracy of the collected road empty parking space detection data set reaches 95.3%, which is 2% higher than the faster RCNN method.
文章引用：梁明秀, 刘秋会, 王林. 一种实时的道路空车位检测算法[J]. 计算机科学与应用, 2020, 10(12): 2439-2446. https://doi.org/10.12677/CSA.2020.1012259

参考文献

[1] Lou, L., Zhang, J., Xiong, Y. and Jin, Y. (2019) An Improved Roadside Parking Space Occupancy Detection Method Based on Magnetic Sensors and Wireless Signal Strength. Sensors, 19, 2348. [Google Scholar] [CrossRef] [PubMed]
[2] Srikanth, S.V., Pramod, P.J., Dileep, K.P., et al. (2009) Design and Im-plementation of a Prototype Smart Parking (SPARK) System Using Wireless Sensor Networks. International Confer-ence on Advanced Information Networking and Applications Workshops, Bradford, 401-406. [Google Scholar] [CrossRef] 
[3] Jo, Y. and Jung, I. (2014) Analysis of Vehicle Detection with WSN-Based Ultrasonic Sensors. Sensors, 14, 14050-14069. [Google Scholar] [CrossRef] [PubMed]
[4] Zhang, Z., Tao, M. and Yuan, H. (2014) A Parking Occupancy Detection Algorithm Based on AMR Sensor. IEEE Sensors Journal, 15, 1261-1269. [Google Scholar] [CrossRef] 
[5] Xiang, X., Lv, N., Zhai, M. and El Saddik, A. (2017) Real-Time Parking Occupancy Detection for Gas Stations Based on Haar-Ada Boosting and CNN. IEEE Sensors Journal, 17, 6360-6367. [Google Scholar] [CrossRef] 
[6] Acharya, D., Yan, W. and Khoshelham, K. (2018) Real-Time Image-Based Parking Occupancy Detection Using Deep Learning. Proceedings of the 5th Annual Re-search@Locate Conference, Adelaide, 9-11 April 2018, 33-40.
[7] Mane, S. and Mangale, S. (2019) Moving Object Detection and Tracking Using Convolutional Neural Networks. 2018 Second International Conference on Intelligent Computing and Control Systems (ICICCS), Madurai, 1809-1813. [Google Scholar] [CrossRef] 
[8] Papageorgiou, C.P., Oren, M. and Poggio, T. (2002) A General Framework for Object Detection. International Conference on Computer Vision, Bombay, 6 August 2002, 555-562.
[9] 安旭骁, 邓洪敏, 史兴宇. 基于迷你卷积神经网络的停车场空车位检测方法[J]. 计算机应用, 2018, 38(4): 935-938.
[10] Saharan, S., Kumar, N. and Bawa, S. (2020) An Efficient Smart Parking Pricing System for Smart City Environment: A Machine-Learning Based Approach. Future Generation Computer Systems, 106, 622-640. [Google Scholar] [CrossRef] 
[11] Ren, S., He, K., Girshick, R., et al. (2017) Faster R-CNN: To-wards Real-Time Object Detection with Region Proposal Networks. IEEE Transactions on Pattern Analysis & Machine Intelligence, 39, 1137-1149. [Google Scholar] [CrossRef] 
[12] Lin, T.Y., Dollár, P., Girshick, R., et al. (2016) Feature Pyra-mid Networks for Object Detection. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Hon-olulu, HI, 936-944. [Google Scholar] [CrossRef] 
[13] Redmon, J. and Farhadi, A. (2017) YOLO9000: Better, Faster, Stronger. IEEE Conference on Computer Vision & Pattern Recognition, Honolulu, HI, 6517-6525. [Google Scholar] [CrossRef]