S.Ia. Krivolapov1, E.D. Solovyov2

1,2 Financial University under the Government of the Russian Federation (Moscow, Russia)
1skrivolapov@fa.ru, 2solzhen@inbox.ru

The problem of determining the brand of a car by its track, which has parameters similar to a certain group of vehicles, is considered. To solve the problem, a machine learning tool is used – a naive Bayesian classifier. To train the classifier, a database containing information about the parameters (track and wheelbase) of 18 car brands is used. It is assumed that based on the results of measurements of the traces left by the vehicle on the ground, approximate values of its parameters will be obtained. Using this information as a test, the trained classifier makes a prediction regarding the make of the car that left traces. The implementation of the described prediction algorithm is performed in Python.

Krivolapov S.Ia., Solovyov E.D. The use of machine learning in the task of conducting automotive examinations. Nonlinear World. 2024. V. 22. № 4. P. 63–69. DOI: https://doi.org/10.18127/ j20700970-202404-08 (In Russian)

1. Burnaev E., Koptelov I., Novikov G., Hanipov T. Avtomatizirovannoe postroenie klassifikatorov na osnove rekurrentnyh nejronnyh setej dlya detekcii proezdov transportnyh sredstv. Sbornik trudov 40-j mezhdisciplinarnoj shkoly-konferencii IPPI RAN. 2016 (In Russian).
2. Koptelov I., Grigor'ev A., Hanipov T., Emel'yanov S., Nikolaev D. Maket avtomaticheskogo klassifikatora transportnyh sredstv. Sbornik trudov 38-j mezhdisciplinarnoj shkoly-konferencii IPPI RAN. 2014. № 2 (In Russian).
3. Yudin D.A., Knysh A.S., Kapustina E.O. Obnaruzhenie avtomobilej na staticheskih izobrazheniyah s ispol'zovaniem metoda ViolyDzhonsa. Sbornik materialov III mezhdunar. nauch.-prakt. konf. «Innovacionnoe razvitie avtomatizacii, informacionnyh i energosberegayushchih tekhnologij, metallurgii i metallovedeniya. Sovremennoe sostoyanie, problematika i perspektivy». M.: MISiS. 2015. S. 280–287 (In Russian).
4. Yudin D.A., Gorshkova N.D., Knysh A.S., Frolov S.V. Raspoznavanie transportnyh sredstv i registraciya ih traektorii dvizheniya na posledovatel'nosti izobrazhenij. Vestnik BGTU im. V.G. Shuhova. 2016. № 6. S. 139–147 (In Russian).
5. Ohn-Bar E., Trivedi M.M. Go with the flow: Improving Multi-View vehicle detection with motion cues. International Conference on Pattern Recognition. 2014. DOI 10.1109/ICPR.2014.709.
6. Rezaei M., Terauchi M. Vehicle Detection Based on Multi‐feature Clues and Dempster‐Shafer Fusion Theory. Conference: 6th Pacific-Rim Symposium on Image and Video Technology, PSIVT 2013. DOI:10.1007/978-3-642-53842-1_6.
7. Viola P., Jones M.J. Rapid object detection using a boosted cascade of simple features. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR’01). 2001. V. 1. S. I-511–I-518.
8. Koshkin A.A. Klassifikaciya sledov dorozhno-transportnyh proisshestvij. Novainfo. 2018. № 80. S. 65–67 (In Russian).
9. Maksimov N.V. Znachimost' sledov pri rassledovanii ekologicheskih prestuplenij. Nacional'naya bezopasnost' v ekologicheskoj sfere: problemy teorii i praktiki. Sbornik materialov Mezhdunarodnoj nauchno-prakticheskoj konferencii. 2017. S. 543–546
   (In Russian).