Detection and classification of traffic objects using the environment digits

Authors

  • R.N. Kvietnyi Vinnitsa National Technical University
  • R.V. Maslii Vinnitsa National Technical University
  • O.M. Kyrylenko Vinnitsa National Technical University

DOI:

https://doi.org/10.31649/1681-7893-2020-39-1-14-20

Keywords:

deep learning, object detection, object classification, DetectNet, DIGITS, KITTI

Abstract

An overview of the architecture of the DetectNet neural network was conducted to study the model of detection and classification of traffic objects. In this case, the structure of the neural network and the format of the input data are considered. The modeling is done using the DIGITS environment. The quality of the model was tested on the image validation dataset KITTI. The results of studying the model of the neural network are presented. The results obtained compared with existing ones.

Author Biographies

R.N. Kvietnyi, Vinnitsa National Technical University

Ph.D., Professor, Head of AIVT Department

R.V. Maslii, Vinnitsa National Technical University

Ph.D., Associate Professor of AIVT department

O.M. Kyrylenko, Vinnitsa National Technical University

graduate student

References

Nagaraj, S. Edge-based street object detection. / Muthiyan, B., Ravi, S., Menezes, V., Kapoor, K., & Jeon, H. // IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computed, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI). – 2017. – С.1-4.

Roman Kvyetnyy, Roman Maslii, Volodymyr Harmash, Ilona Bogach, Andrzej Kotyra, Żaklin Grądz, Aizhan Zhanpeisova, Nursanat Askarova. Object detection in images with low light condition, Proc. SPIE 10445, Photonics Applications in Astronomy, Communications, Industry, and High Energy Physics Experiments 2017, 104450W (7 August 2017); doi: 10.1117/12.2281001;

Roman Maslii Using Local Binary Patterns for Face Detection in Half-Tone Imagery [Electronic resource] / Maslii Roman // Scientific papers of Vinnytsia National Technical University. - 2008. - No. 4. - [Electronic resource] Access mode: http://praci.vntu.edu.ua/index.php/praci/article/view/95

DetectNet: Deep Neural Network for Object Detection in DIGITS. [Electronic resource] Access mode: https://devblogs.nvidia.com/detectnet-deep-neural-network-object-detection-digits/

Victor Sineglazov Deep neural networks for solving problems of recognition and classification of images [Electronic resource] / Sineglazov V., Chumachenko O. // - Information Technologies and Computer Modeling 2017.– [Electronic resource] Access mode: http://itcm.comp-sc.if.ua/2017/Sineglazov.pdf

Holupka, E. J. The Detection of Implanted Radioactive Seeds On Ultrasound Images Using Convolution Neural Networks. / Rossman, J., Morancy, T., Aronovitz, J., & Kaplan, М. D. // 1st Conference on Medical Imaging with Deep Learning (MIDL 2018). – 2018.

NVIDIA DIGITS [Електронний ресурс] – Режим доступу до ресурсу: https://developer.nvidia.com/digits

Zhang, S. Fcn-rlstm: Deep spatio-temporal neural networks for vehicle counting in city cameras. / Wu, G., Costeira, J. P., & Moura, J. M. // In Computer Vision (ICCV), IEEE International Conference. – 2017. рр. 3687-3696.

Zheng Lou. Dataset bias analysis on autonomous driving [Електронний ресурс] – Режим доступу до ресурсу: https://cs230.stanford.edu/projects_spring_2018/reports/8289902.pdf

Fritsch Jannik. A new performance measure and evaluation benchmark for road detection algorithms. / Tobias Kuehnl, Andreas Geiger. // 16th International IEEE Conference on Intelligent Transportation Systems (ITSC 2013). – 2013.

Downloads

Abstract views: 185

Published

2021-01-08

How to Cite

[1]
R. Kvietnyi, R. Maslii, and O. Kyrylenko, “Detection and classification of traffic objects using the environment digits”, Опт-ел. інф-енерг. техн., vol. 39, no. 1, pp. 14–20, Jan. 2021.

Issue

Section

Systems Of Technical Vision And Artificial Intelligence, Image Processing And Pattern Recognition

Metrics

Downloads

Download data is not yet available.