Rinat SadekovNational University of Science and Technology MISIS · Engineering Cybernetics Research Faculty
Rinat Sadekov
Doctor of Engineering
I develop algorithms for processing inertial, gps, weel and visual sensors for a self-driving car
About
29
Publications
3,974
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
202
Citations
Introduction
Publications
Publications (29)
This article describes an approach to evaluating the localization accuracy of a tram moving in an urban environment. The tram needs to be localized with a submeter level of accuracy. Since GPS information in urban environment does not provide such a level of accuracy, a solution based on the use of vision system information is proposed. By using ke...
The article analyzes the works devoted to outdoor navigation of unmanned aerial vehicles
(UAVs) in GNSS-denied environments using computer vision systems (optical range cameras). The algorithms addressed are based on matching of UAV-generated images with the
available georeferenced terrain images. Images are matched either pixel by pixel, by their...
In this paper we propose a solution to the problem of predicting the behavior of vehicles on the road scene based on image analysis. A 3D detection deep neural network is used to detect objects and estimate their position. The kinematic bicycle model is considered as a behavior model, with each vehicle being processed separately, i.e. without takin...
The article proposes an approach to recognize dishes located on a tray to make it possible to automatically calculate the cost of an order. The solution to the problem has several stages. At the first stage, using the Yolo deep learning network, the areas of the arrangement of dishes on the tray are determined, at the second, menu items are specifi...
The paper considers an original autonomous correction algorithm for UAV navigation system based on comparison between terrain images obtained by onboard machine vision system and vector topographic map images. Comparison is performed by calculating the homography of vision system images segmented using the convolutional neural network and the vecto...
The paper considers an original autonomous correction algorithm for UAV navigation system based on comparison between terrain images obtained by onboard machine vision system and vector topographic map images. Comparison is performed by calculating the homography of vision system images segmented using the convolutional neural network and the vecto...
The paper proposes a solution to the automatic along the straw rows by means of the images from the camera, installed in the cab of the harvester. The U-Net is used to recognize straw rows in the image. The edges of the row are approximated in the segmented image by the curved lines and further converted into the harvester coordinate system for the...
An approach is proposed to reduce systematic error that arises in nonlinear problems of estimating parameters due to inaccurate selection of the boundaries of intervals for constructing a grid. It is shown that the choice of boundaries is hampered by the small number of available measurements and their low accuracy. An increase in the accuracy of t...
We focus on car navigation method using vision system and digital road map. Vision system installed onboard the car serves as an additional source of navigation information. It is used to determine the car lateral deviation relative to the road axial line, range and bearing to the intersection center. Algorithm generating these parameters is descri...
Computer vision hardware and software for navigation and mapping are discussed. The system includes two video cameras. The main features of the software used for calculation and visualization of the library are described in sufficient detail. A method for calibration of a computer vision system using a pair of images with several calibration patter...
Results of a study to create a high-precision system for determining astronomical azimuth are presented. The composition, structural features, and algorithmic support of the device are considered. Results of an estimation of the precision of azimuthal measurements that confirm the high precision characteristics of the device are presented.
The limits used in the recognition of an arrival from a road trip in the problem of complex processing of the information of automobile navigation systems and digital road maps where there are no constraints on the movement of the automobile are evaluated. It is shown that the precision of estimators in the determination of navigation parameters ca...