應用深度學習MobileNets SSD偵測真實環境中多樣的汽車車牌

Diversity Vehicle License Plate Detection in the Real Environment via Deep Learning MobileNets SSD

蔡俊明(Chun-Ming Tsai) ； 林俞佑(Yu-You Lin)

《國教新知》 65卷4期 (2018/12) Pp. 83-97

https://doi.org/10.6701/TEEJ.201812_65(4).0006

摘要

路邊停車收費員每天在路上騎電動摩托車查看停車格是否有停車，若雨刷下沒有夾繳費單，要用PDA輸入車牌號碼和停車格號碼，再列印繳費單，並把繳費單夾在雨刷下，再用數位相機拍照。若是雨刷下已經有繳費單，要檢查和蓋章，這實在是一件很辛苦和危險的工作。若是能有一套智慧型車牌偵測系統，停車收費員戴在頭上，就像戴眼鏡一樣，停車收費員只要騎著電動機車，慢慢從路邊停車格旁邊開過去，用眼睛往路邊停車格內車輛一看，此系統就可以自動把車牌偵測出來。然而在真實環境中的車牌具多樣性，例如不同拍攝角度，不同光源，甚至車牌背景顏色與車身顏色一樣。為了解決上述車牌偵測問題，我們利用深度學習的Mobile NetsSSD來訓練和偵測車牌，經實驗在234張上述真實且多樣的車牌，所用的方法，整體的準確率為93.59%。

關鍵字

深度學習 ； 車牌偵測 ； 路邊停車收費員

並列摘要

Roadside parking toll collectors ride electric motorcycles on the road every day to check whether there is parking in the parking space. If there is no payment slip under the wiper, use the PDA to enter the license plate number and the parking space number, print the payment slip and place under the wiper, take a photo with a digital camera. If there is already a payment slip under the wiper, it should be checked and be stamped. This job is very difficult and dangerous. If there is an intelligence license plate detection system, the parking toll collector wears it on the head, just like wearing glasses. The parking toll collector just rides the electric motor car and slowly drives away from the roadside parking space, using the eyes to the roadside parking space. Once the vehicle is inside, the system can automatically detect the license plate. However, the license plates in the real environment are diverse, such as different shooting angles, different light sources, and even the license plate background is the same as the vehicle body color. In order to solve the above-mentioned license plate detection problem, we use the deep learning MobileNets SSD to train and detect the license plates. The experiments detect in 234 real and diverse license plates. The overall of accuracy detection rate is 93.59%.

並列關鍵字

deep learning ； vehicle license plate detection ； roadside parking toll collector

參考文獻

Anagnostopoulos, C.-N. E.,Anagnostopoulos, I. E.,Psoroulas, I.D.,Loumos, V.,Kayafas, E.(2008).License plate recognition from still images and video sequences: A survey.IEEE Transactions on Intelligent Transportation Systems.9(3),377-391.

Ashtari, A. H.,Nordin, M. J.,Fathy, M.(2014).An Iranian license plate recognition system based on color features.IEEE Transactions on Intelligent Transportation Systems.14(4),1690-1705.

Du, S.,Ibrahim, M.,Shehata, M.,Badawy, W.(2013).Automatic license plate recognition (ALPR): A state-of-the-art review.IEEE Transactions on Circuits and Systems for Video Technology.23(2),311-325.

Howard, A.G., Zhu, M.L., Chen, B., Kalenichenko, D., Wang, W.J., Weyand, T., Andreetto, M., and Adam, H. (2017). MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. In arXiv.1704.04861v1 [cs.CV] 17 Apr 2017.

Laroca, R.,Severo, E.,Zanlorensi, L.A.,Oliveira, L.S.,Goncalves, G.R.,Schwartz, W.R.,Menotti, D.(2018).A Robust Real-Time Automatic License Plate Recognition Based on the YOLO Detector.International Joint Conference on Neural Networks (IJCNN).(International Joint Conference on Neural Networks (IJCNN)).