Registration Number
0218U007941,
0118U005366
, R & D reports
Title
Development of methods for detecting unmanned aerial vehicles on the sequence of video images
popup.stage_title
Head
Neuimin Oleksandr Stanislavovych,
Registration Date
27-12-2018
Organization
National Technscal University of Ukraine "Kiev Polytechnic Institute".
popup.description2
The development of technology for building unmanned aerial vehicles (UAVs), their accessibility and mass use for solving a wide range of tasks has led to the emergence of a new class of threats: terrorist use, photographing secret objects, invading a forbidden territory, etc. Therefore, applied scientific work is which aims to develop methods for detecting UAVs on video sequences. Detection of moving objects should occur at maximum range when they fall into the field of view of the optical system. Under these conditions, the object in each frame is considered as a point. Detecting a moving point object in an image is quite a challenge, since the intensity of the object in the image is low due to the low signal-to-noise ratio (SNR). To increase the efficiency of tracking objects, decisive statistics (DS) of marks are used. Widespread for optimal detection of the trajectory of the object is consistent with the Wald criterion, which allows to obtain a significant gain in time with given requirements for decision-making probabilities. However, it loses its optimal properties if the actual SNR of the object is lower than the specified one. Therefore, of practical importance is the task of identifying the trajectory of an object using DS of marks with an unknown SNR. In this paper, the actual scientific problem was solved: development of methods for detecting a point object using DS of marks, which allow increasing the efficiency of its detection with small SNRs and do not require significant computational costs.
Product Description
popup.authors
Неуймін Олександр Станіславович
Товкач Ігор Олегович
popup.nrat_date
2020-04-02
