TRACKING A SMALL MOVING TARGET IN A VIDEO SEQUENCE AMONG MANY SIMILAR OBJECTS
A.V. Likhachev
Institute of Automation and Electrometry, Siberian Branch, Russian Academy of Sciences, Novosibirsk, Russia
Keywords: searching for the target’s trajectory using a video sequence, a multitude of small objects in the image, cloudy background
Abstract
An algorithm has been developed that solves the problem of searching for an object moving in the plane of a video frame along a curve of a given shape among many similar objects in a video sequence. It is based on iterating through possible trajectories defined by the positions of the objects in two selected frames. For each trajectory, the initial speed of the target and its acceleration are estimated. The trajectory is selected such that, while moving along it, the target passes closest to the positions of the objects recorded in all frames. The algorithm has been tested on model data. Dependencies of its efficiency on the length of the video sequence, the average number of objects in the image, the shape of the trajectory, the nature of the movement, and the noise level have been obtained.
|
