The paper considers aspects of foreground detection in dynamic scenes of intelligent transport system based on computer vision and artificial intelligence. Traditional and recent background modeling models have been considered. Nonparametric approach for background subtraction based on the kernel model was used as the most appropriate in dynamic environment. The value of the estimated kernel density function for each pixel of original image was compared with threshold value, estimated by Otsu’s method. The proposed kernel density estimation method was verified on video-stream containing moving objects and indicated good performance for Unmanned Aerial Vehicles application.
FOREGROUND DETECTION IN DYNAMIC SCENES OF INTELLIGENT TRANSPORT SYSTEMS
2019
Aufsatz (Zeitschrift)
Elektronische Ressource
Unbekannt
Metadata by DOAJ is licensed under CC BY-SA 1.0
Special issue on background modeling for foreground detection in real-world dynamic scenes
British Library Online Contents | 2014
|Learning Complementary Saliency Priors for Foreground Object Segmentation in Complex Scenes
British Library Online Contents | 2015
|Counting vehicles in urban traffic scenes using foreground time-spatial images
IET | 2017
|