Detection of foreign objects at railway crossings (RC) is an important task from the point of view of ensuring the safety of train traffic. Railway traffic management systems need reliable means of monitoring and determining the state of the RC. The use of digital surveillance cameras placed on the RC and the analysis of video information obtained with their help increases the efficiency and speed of response to dangerous situations. This article presents a method for processing video data using object detection algorithms based on key points to determine the state of the RC. The proposed approach has been tested on the basis of video materials obtained from surveillance cameras installed at RC facilities. Within the framework of computer modeling, the results of the application of the proposed method are obtained, taking into account the assessment of information indicators of the quality of the train safety system
railway crossing, obstacle, contour, key points, object classification, train traffic control
1. Tastimur C., Karakose M., Akın E. Image Processing Based Level Crossing Detection and Foreign Objects Recognition Approach in Railways // International Journal of Applied Mathematics, Electronics and Computers. - 2017. - Special Issue. - rr. 19-23.
2. Pu Y.R., Chen L.W, Lee S. H. Study of moving obstacle detection at railway crossing by machine vision // Informational Technology Journal. - 2014. - Vol. 13.16. - pp 2611-2618.
3. Fahmi Sh.S., Shatalova N.V., Visloguzov V.V. Kostikova E.V. Mnogoprocessornaya transportnaya videosistema nablyudeniya na baze tehnologii «sistema na kristalle» // Morskie intellektual'nye tehnologii. - 2020. - № 3-1 (49). - S. 172-180.
4. Fahmi Sh.S., Shatalova N.V., Kostikova E.V., Kolesnichenko S.V. Mnogoprocessornaya sistema obnaruzheniya i slezheniya za morskimi sudami na baze tehnologii «sistema na kristalle» // Morskie intellektual'nye tehnologii. -2021. -T.1, № 2 (53). - S. 165-172.
5. Ivanov A.V., Fahmi Sh.S. Obrabotka videoinformacii v transportnyh videosistemah real'nogo vremeni: monografiya. - SPb.: Izdatel'sko-poligraficheskaya associaciya VUZ, 2021. - 222 s.
6. Fahmi, Sh. S. Metod i algoritmy obnaruzheniya i raspoznavaniya dorozhnyh znakov/ Sh. S. Fahmi, E. V. Kostikova, M.S. Kryukova, I. A. Gavrilov, K.M. Zaydulin // Voprosy radioelektroniki, ser. Tehnika televideniya. - 2018. - Vyp. 3. - S. 95-101.
