Real-time multiple object tracking algorithm for adaptive tra-c control systems : доклад, тезисы доклада

Описание

Тип публикации: доклад, тезисы доклада, статья из сборника материалов конференций

Конференция: Applied Methods of Statistical Analysis. Statistical Computation and Simulation - AMSA'2019; Novosibirsk; Novosibirsk

Год издания: 2019

Ключевые слова: real-time tracking, object recognition, convolutional neural network, distance metrics, machine learning, computer vision, adaptive tra-c control systems

Аннотация: This study is dedicated to development of a tracking algorithm. The main objective for such algorithm is its mandatory implementation with minimal computing with minimal computing resources, including the provision of a data transmission channel. The requirement is dictated by the need to obtain information about the ratio of vehicПоказать полностьюles passing through the intersection in real time. The data is necessary for adaptive correction of the time shifts for coordination plans of tra-c lights. The development of the proposed method assumes the classi cation of the received tracks for the recognition of emergency situations.

Ссылки на полный текст

Издание

Журнал: Applied Methods of Statistical Analysis. Statistical Computation and Simulation - AMSA'2019

Номера страниц: 530-536

Место издания: Novosibirsk

Издатель: Новосибирский государственный технический университет

Персоны

  • Kovalenko Mikhail A. (Siberian Federal University)
  • Sergeeva Natalia A. (LLC Rd-science)

Вхождение в базы данных