臺灣博碩士論文加值系統

近年來，由於自動化停車管理系統的高便利性和高效率性，使得自動停車管理系統在現代城市地區變得非常受歡迎。除此之外，車輛計數比起其他自動化系統更早被應用於智慧停車場的管理系統中，因此我們致力於開發一種基於立體計算機視覺的新型車輛計數機制。在此篇論文中，我們首先使用基於場景流的方法建立深度圖，我們也修改了sigmoid function來改變所獲得深度圖的分布，並且設計用來估計深度圖中的視差值的標定版樣式。因此，我們提出了一種藉由改變視差值分布的深度影像的車輛計數機制，可以很容易地區分是否有車輛。並且，透過我們提出的車輛偵測和計數的方法，監控攝像機可以確保車輛接近距離入口適當的位置。在系統評估的部分，我們在室內停車場環境中錄製了六天由早上8點至下午7的影片約66小時，而實驗結果顯示，我們的方法在車輛計數中有不錯的表現，Precision rate 和 Recall rate分別達到99.56％和98.29％。由此可知，我們的方法可以避免交通高峰期的車輛計數錯誤。未來，更希望能完全取代普遍安裝在停車場中的車輛計數機制的地下感應線圈。

In recent years, automated parking management systems have become very popular in modern urban areas owing to for their convenience and efficiency. As vehicle counting plays an important role in other applications, so we endeavored to develop a novel mechanism for counting vehicles, based on stereo computer vision. In this work, we first establish depth maps of a given image using a scene flow-based approach. Then, we designed a modified sigmoid function to change the histogram distribution in the obtained depth maps, through the disparity threshold estimated from our self-designed disparity calibration board. As a result, we proposed a vehicle counting mechanism using by the changed disparity histogram that can easily distinguish whether there is a presence of vehicle or not. Consequently, through our proposed vehicle detection and counting method, a surveillance camera can make sure if a vehicle is approaching an entrance, and it can take a clear snapshot of the license plate for automatic recognition. The proposed system was evaluated on six sets of videos of about 66 hours recorded in an indoor parking environment. The experimental results show that our method exhibits good performance in vehicle counting. The Precision rate and Recall rate reach 99.56% and 98.29%, respectively. Our method can avoid errors in vehicle counting during peak periods of traffic. In the future, the induction underground coil in a parking lot can be fully replaced by the proposed vehicle counting mechanism.

1 Introduction
1.1 Motivation ................... 1
1.2 Objective .................... 3
1.3 Thesis Overview ......... 4

2 Related Literature
2.1 Applications for Smart Parking Management ...... 5
2.2 Vehicle Detection ................................................ 7
2.3 Establish the Depth Map ..................................... 8

3 The Vehicle Counting System
3.1 The calibration of two cameras ............................................................... 11
3.2 Establish depth maps using growing correspondence of scene flow ..... 16
3.3 The Method of vehicles counting ........................................................... 22

4 Experiment and Results
4.1 Experiment on Disparity Calibration Board ......... 33
4.2 Experimental Results of Vehicle Counting .......... 36
4.3 Main Issues of Analysis Results .......................... 41

5 Conclusion and Future Work
5.1 Conclusion ........... 44
5.2 Future Work ......... 45

[1] : Ministry of transportation and communications of the republic of china. http: //www.thb.gov.tw/ (2017)
[2] Mimbela, L.E.Y., Klein, L.A.: Summary of vehicle detection and surveillance technolo- gies used in intelligent transportation systems. (2000)
[3] Idris, M., Leng, Y., Tamil, E., Noor, N., Razak, Z.: park system: a review of smart parking system and its technology. Inf. Technol. J 8 (2009) 101–113
[4] Geng, Y., Cassandras, C.G.: New“smart parking”system based on resource allocation and reservations. IEEE Transactions on Intelligent Transportation Systems 14 (2013) 1129–1139
[5] Wang, H., He, W.: A reservation-based smart parking system. In: IEEE Conference on Computer Communications Workshops, INFOCOM WKSHPS, IEEE (2011) 690–695
[6] Swathy, M., Nirmala, P., Geethu, P.: Survey on vehicle detection and tracking tech- niques in video surveillance. International Journal of Computer Applications 160 (2017)
[7] Chintalacheruvu, N., Muthukumar, V., et al.: Video based vehicle detection and its application in intelligent transportation systems. Journal of transportation technologies 2 (2012) 305
[8] Xia, Y., Shi, X., Song, G., Geng, Q., Liu, Y.: Towards improving quality of video- based vehicle counting method for tra c ow estimation. Signal Processing 120 (2016) 672–681
[9] Wen, X., Shao, L., Xue, Y., Fang, W.: A rapid learning algorithm for vehicle classi - cation. Information Sciences 295 (2015) 395–406
[10] Zhan, W., Ji, X.: Algorithm research on moving vehicles detection. Procedia Engineer- ing 15 (2011) 5483–5487
[11] Rashidi, A., Fathi, H., Brilakis, I.: Innovative stereo vision-based approach to generate dense depth map of transportation infrastructure. Transportation Research Record: Journal of the Transportation Research Board (2011) 93–99
[12] Marr, D., Poggio, T.: A computational theory of human stereo vision. Proceedings of the Royal Society of London B: Biological Sciences 204 (1979) 301–328
[13] Hirschmuller, H.: Accurate and e cient stereo processing by semi-global matching and mutual information. In: IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR. Volume 2., IEEE (2005) 807–814
[14] Gehrig, S.K., Eberli, F., Meyer, T.: A real-time low-power stereo vision engine us- ing semi-global matching. In: International Conference on Computer Vision Systems, Springer (2009) 134–143
[15] Banz, C., Hesselbarth, S., Flatt, H., Blume, H., Pirsch, P.: Real-time stereo vi- sion system using semi-global matching disparity estimation: Architecture and fpga- implementation. In: International Conference on Embedded Computer Systems, SAMOS, IEEE (2010) 93–101
[16] Hirschmuller, H.: Stereo processing by semiglobal matching and mutual information. IEEE Transactions on pattern analysis and machine intelligence 30 (2008) 328–341
[17] Cech, J., Sara, R.: E cient sampling of disparity space for fast and accurate match- ing. In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR, IEEE (2007) 1–8
[18] Čech, J., Sanchez-Riera, J., Horaud, R.: Scene ow estimation by growing correspon- dence seeds. In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR, IEEE (2011) 3129–3136
[19] Hung, C.H., Xu, L., Jia, J.: Consistent binocular depth and scene ow with chained temporal pro les. International journal of computer vision 102 (2013) 271–292
[20] Liu, F., Philomin, V.: Disparity estimation in stereo sequences using scene ow. In: BMVC. Volume 1. (2009) 2
[21] Vedula, S., Baker, S., Rander, P., Collins, R., Kanade, T.: Three-dimensional scene ow. In: The Proceedings of the Seventh IEEE International Conference on Computer Vision. Volume 2., IEEE (1999) 722–729
[22] Yan, Z., Xiang, X.: Scene ow estimation: A survey. arXiv preprint arXiv:1612.02590 (2016)
[23] Bouguet, J.Y.: Pyramidal implementation of the a ne lucas kanade feature tracker description of the algorithm. Intel Corporation 5 (2001) 4
[24] Harris, C., Stephens, M.: A combined corner and edge detector. In: Alvey vision conference. Volume 15., Manchester, UK (1988) 10–5244
[25] Zhang, Z.: A exible new technique for camera calibration. IEEE Transactions on pattern analysis and machine intelligence 22 (2000) 1330–1334
[26] : Kinect - wikipedia. https://zh.wikipedia.org/wiki/Kinect (2017)
[27] Lucas, B.D., Kanade, T., et al.: An iterative image registration technique with an
application to stereo vision. (1981)
[28] Han, J., Moraga, C.: The in uence of the sigmoid function parameters on the speed of backpropagation learning. From Natural to Arti cial Neural Computation (1995) 195–201