Analysis of Recorded Speeds from LRT Tachographs Using Image Processing
Lucas Gouveia Omena Lopes1; José Wellington Pedro da Silva Filho2
1 Universidade Federal de Alagoas; 2 CBTU-MACEIÓ
doi:10.20906/CPS/CILAMCE2017-0152
Resumo
This paper focus in the analysis of recorded speeds from LRT tachographs using image processing. Tachographs are mandatory equipment for commercial vehicles that carry loads and passengers. The main function of this tool is to serve as data source for transportation companies and inspection agencies in case of possible infractions. The functions and the data collected by each device may change according to task that It will perform, being more common the speed and time registration. Traffic Engineers responsible for the operation of a Light Rail Transit (LRT) use this information for inspection, planning of the trips and elaboration of technical reports. However, manufacturers do not develop effective tools for practical and accurate interpretation of the measured speeds, making this task costly and subject to error. In order to attend this demand, this paper proposes a methodology for the analysis of tachograph velocity data in LRTs using image processing. Techniques of image processing and computer vision have become popular over the years. It is possible to extract information from the smallest units of the images through these techniques. The strategy consists in the extraction of the corresponding velocity and time values from a differentiation in the intensity of the pixels, making use of the images from the data collected by the tachographs. It is possible to determine, from interpolations, the functions that represent the speed and position of the vehicle along the trip through the speed and time values. The methodology presented has the advantage of being an automated software and having an adjustable precision, since this depends only on the quality of the image. Calibration tests are performed for the strategy using travel data recorded by the tachographs of LRTs. These tests are also used as a base to improve the user interface of the software. The results show a good correspondence between the values measured through the methodology currently used by the traffic engineers and the strategy presented i
Palavras-chave: Tachographs ; Image Processing; Light Rail Transit