The week's pick
Random Articles
Reseach Article
An Improved Frame Difference Background Subtraction Technique for Enhancing Road Safety at Night
| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 183 - Number 1 |
| Year of Publication: 2021 |
| Authors: Clement Alabi, James Ben Hayfron-Acquah |
10.5120/ijca2021921280
|
Clement Alabi, James Ben Hayfron-Acquah . An Improved Frame Difference Background Subtraction Technique for Enhancing Road Safety at Night. International Journal of Computer Applications. 183, 1 ( May 2021), 38-45. DOI=10.5120/ijca2021921280
Abstract
Night vision technology allows for vision in darkness as though just as in broad daylight. To this effect, this research is carried out to see the possibility of using thermal-based night vision to curb road accidents that have become rampant on the roads and consequently reduce damage and the loss of life. Thermal sensors have a higher range than the average vehicle head and taillights, which use the visible light spectrum. This study proposes a concept where signals are received from a FLIR camera, the grayscale images are processed by an enhanced technique; the WBODY_DET () algorithm in the MATLAB environment. This algorithm identifies and highlights all hot objects by drawing colored boundaries around them. The algorithm is configurable and can highlight categories of objects in a scene. For instance, engines can be highlighted with different colors from living bodies like humans, cattle, sheep, goats, cats, dogs, etc. for easy recognition by drivers/observers for further response. The crux of this algorithm is an object detection technique that uses simple thresholding. The accuracy of this algorithm is enhanced by using morphological operators and the property of connectedness within the resulting binary image to remove false positives. This heat-based vision system is best suited for night driving since all moving objects emit some form of heat energy due to the work involved in the movement. Findings show that the proposed model is highly efficient since it is fast, has over a 90% detection accuracy and it allows for easy identification of danger on the roads at night from longer distances and under bad visibility levels. It is recommended that all vehicles should be upgraded with this technology so the roads will be safer. It is also recommended that long-range cameras should be used as the longer the range of the camera the longer the reaction times the system will provide drivers and prevent a collision.
References
- Babaee, M., Dinh D, T., and Rigoll, G. A deep convolutional neural network for video sequence background subtraction. Pattern Recognition Volume 76, April 2018, 635-649. https://doi.org/10.1016/j.patcog.2017.09.040.
- Benton, S. Background subtraction, part 1. MATLAB models: EDN Network., 2010.
- Cleverpedia. What is Grayscale Coloring? 2020. http://www.cleverpedia.com/grayscale-coloring-tutorial/
- Cui, B., and Créput, J-C. Moving Object Detection and Tracking Based on Three-Frame Difference and Background Subtraction with Laplace Filter. International Conference on Artificial Intelligence and Soft Computing ICAISC - P. 2007 Modeling and Simulation Design, 2018, pp 3-13.
- GHANAWEB. Ghana records more than 2000 road accident deaths. 2020.https://www.ghanaweb.com/GhanaHomePage/NewsArchive/Ghana-records-more-than-2000-road-accident-deaths-1120157
- Gonzalez, R.C., Woods, R.E. (2002). Digital Image Processing. Prentice Hall, 2002. https://doi.org/10.1016/j.compeleceng.2013.03.020.
- Lee, E. Yuxunion-night vision thermal imaging (laser infrared camera)-English Version, (May,2018). https://youtu.be/GtPYLSw168s.
- Liu, Q., Zhuang. and Kong, S. Detection of pedestrians for far-infrared automotive night vision systems using learning-based method and head validation. IOP Publishing Ltd. Journal of Sensors Volume 2016, ArticleID3403451, pp 8, 2013. http://dx.doi.org/10.1155/2016/3403451
- NRSC. WELCOME TO HOMEPAGE OF NRSC, (2018). http://nrsa.gov.gh/index.php/statistics.
- Shahbaz, A., Hariyono, J. and Jo, K.H. Evaluation of background subtraction algorithms for video surveillance. Frontiers of Computer Vision (FCV),21st Korea-Japan Joint Workshop, 10.1109, 2015, FCV.7103699.
- Singla, N. Motion Detection Based on Frame Difference Method. International Journal of Information & Computation Technology. ISSN 0974-2239 Volume, 2014, 4(15) pp. 1559-1565
- Sobral, A., Javed, S., Jung, S.K, Bouwmans, T. and Zahzah, E. Online Stochastic Tensor Decomposition for Background Subtraction in Multispectral Video Sequences. ICV Workshop, Computer vision Foundation, 2015.http://openaccess.thecvf.com/content_iccv_2015_workshops/w24/papers/Sobral _Online_Stochastic_Tensor_ICCV_2015_paper.pdf
- Tekalp, A.M. Digital Video Processing. 2nd edition, Prentice Hall Press, 2015b, Upper Saddle River, NJ, USA. ISBN:0133991008 9780133991000.
- Zhang, R., and Ding, J. Object Tracking and Detecting Based on Adaptive Background Subtraction. Procedia Engineering, 29, 2012, 1351-1355, https://doi.org/10.1016/j.proeng.2012.01.139.
Index Terms
Keywords