Journal of Convergence for Information Technology (융합정보논문지)

Convergence Society for SMB (중소기업융합학회)
권호 표지
DOI QR코드

DOI QR Code

Classification of Trucks using Convolutional Neural Network

합성곱 신경망을 사용한 화물차의 차종분류

  • Lee, Dong-Gyu (Division of IT Convergence Engineering, Shinhan University)
  • 이동규 (신한대학교 IT융합공학부)
  • Received : 2018.10.01
  • Accepted : 2018.12.20
  • Published : 2018.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

This paper proposes a classification method using the Convolutional Neural Network(CNN) which can obtain the type of trucks from the input image without the feature extraction step. To automatically classify vehicle images according to the type of truck cargo box, the top view images of the vehicle are used as input image and we design the structure of the CNN suitable for the input images. Learning images and correct output results is generated and the weights of neural network are obtained through the learning process. The actual image is input to the CNN and the output of the CNN is calculated. The classification performance is evaluated through comparison CNN output with actual vehicle types. Experimental results show that vehicle images could be classified with more than 90 percent accuracy according to the type of cargo box and this method can be used for pre-classification for inspecting loading defect.

본 논문에서는 화물차 차종을 분류하기 위해서 특징추출단계 없이 입력영상으로부터 차종분류결과를 얻을 수 있는 합성곱 신경망을 사용한 분류방법을 제안한다. 차량의 위에서 촬영된 영상을 입력으로 사용하고 입력영상에 적합한 합성곱 신경망의 구조를 설계한다. 차종과 화물칸의 형태에 따라 차종을 자동 분류하기 위한 학습데이터를 생성하고 지도학습의 형태로 학습시키기 위해 분류된 영상과 올바른 출력결과를 제시하여 신경망의 가중치를 학습시킨다. 실제 영상을 입력하여 합성곱 신경망의 출력을 계산하였고 실제 차종과의 비교를 통해 분류 성능을 평가 하였다. 실험결과 화물의 차종과 적재함의 형태에 따라 90%이상의 정확도로 영상을 분류할 수 있었고, 적재불량 검사의 사전 분류에 활용될 수 있다.

Keywords

JKOHBZ_2018_v8n6_375_f0001.png 이미지

Fig. 1. Convolution operation

JKOHBZ_2018_v8n6_375_f0002.png 이미지

Fig. 2. 3x3 MaxPooling(stride=2, shadow region=pooling center)

JKOHBZ_2018_v8n6_375_f0003.png 이미지

Fig. 3. Architecture of proposed CNN

JKOHBZ_2018_v8n6_375_f0004.png 이미지

Fig. 4. the example of learning data (a) background, output=0 (b) car, output=1 (c) small truck, output=2 (d) large truck, output=3 (e) small van, output=4 (f) large van, output=5

JKOHBZ_2018_v8n6_375_f0005.png 이미지

Fig. 5. (a) Input Image (b) Output value of softmax (c) 6 feature map of Conv1

Table 1. Classification category and learning data

JKOHBZ_2018_v8n6_375_t0001.png 이미지

Table 2. Classification result

JKOHBZ_2018_v8n6_375_t0002.png 이미지

References

  1. R. Girshick, J. Donahue, T. Darrell & J. Malik..(2014). Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation. Proceedings of CVPR 2014, 580-587.
  2. G. S. Kim. (2002). Morphological Vehicle Classification Algorithm for Intelligent Transportation System. Journal of Korea Multimedia Society, 5(1), 10-17.
  3. W. S. Shin & C. H. Lee. (2008). Efficient Learning and Classification for Vehicle Type using Moving Cast Shadow Elimination in Vehicle Surveillance Video. The KIPS Transactions, 15B(1), 1-8. https://doi.org/10.3745/KIPSTB.2008.15-B.1.1
  4. Q. Paul, J. H. Hyeon & S. W. Ha. (2017). A Study on Attitude Estimation of UAV Using Image Processing. Journal of Convergence for Information Technology, 7(5), 137-148. https://doi.org/10.22156/CS4SMB.2017.7.5.137
  5. H. Huttunen, F. S. Yancheshmeh & K. Chen. (2016). Car type recognition with Deep Neural Networks. Proceedings of the IEEE Intelligent Vehicles Symposium, 1016-1021. DOI : 10.1109/IVS.2016.7535529
  6. Y. O. Adu-Gyamfi & A. Sharma. (2017). Automated Vehicle Recognition with Deep Convolutional Neural Networks. Transportation Research Record Journal of the Transportation Research Board, 2645, 113-122. DOI : 10.3141/2645-13
  7. J. C. Han, B. C. Koo & K. J. Cheoi. (2017). Obstacle Detection and Recognition System for Autonomous Driving Vehicle. Journal of Convergence for Information Technology, 7(6), 229-235. https://doi.org/10.22156/CS4SMB.2017.7.6.229
  8. B. Su, J. Shao, J. Zhou, X. Zhang, L. Mei & C. HuY. (2016). The Precise Vehicle Retrieval in Traffic Surveillance with Deep Convolutional Neural Netwoks. International Journal of Information and Electronics Engine, 6(3), 192-197. https://doi.org/10.18178/IJIEE.2016.6.3.622
  9. Z. Dong, Y. Wu, M. Pei & Y. Jia. (2015). Vehicle Type Classification Using a Semisupervised Convolutional Neural Network. IEEE Transactions on Intelligent Transportation systems, 16(4), 2247-2256. DOI : 10.1109/TITS.2015.2402438
  10. P. K. Kim & K. T. Lim. (2017). Vehicle Type Classification Using Bagging and Convolutional Neural Network on Multi View Surveillance Image. IEEE Conference on CVPRW, 11(2), 41-46. DOI : 10.1109/CVPRW.2017.126
  11. E. Rhee. (2017). Security Algorithm for Vehicle Type Recognition. Journal of Convergence for Information Technology, 7(2), 77-82. https://doi.org/10.22156/CS4SMB.2017.7.2.077
  12. W. K. Park, Y. Choi, H. K. KIM, G. S. Choi & H. Y. Jung. (2017). Night-time Vehicle Detection Method Using Convolutional Neural Network. The Korean Journal of IEMEK, 12(2), 113-120.
  13. M. K. Kwon & H. S. Yang. (2017). A scene search method based on principal character identification using convolutional neural network. Journal of Convergence for Information Technology, 7(2), 31-36. https://doi.org/10.22156/CS4SMB.2017.7.2.031
  14. Peemen, Mesman, Bart, Corporaal & Henk.. (2011). Speed sign detection and recognition by convolutional neural networks. Proceedings of the 8th International Automotive Congress, 11(2), 162-170.
  15. J. H. Ku. (2018). A Study on Adaptive Learning Model for Performance Improvement of Stream Analytics. Journal of Convergence for Information Technology, 8(1), 201-206. https://doi.org/10.22156/CS4SMB.2018.8.1.201