The research of multi-object tracking algorithm using Kalman filtering method. (2nd September 2019)

• Record Type: Journal Article
• Authors: Liu, Shuqing
• Is Part Of: International journal of innovative computing and applications. Volume 10:Number 2(2019)
• Page Start: 107
• Page End: 114
• Publication Date: 2019-09-02
• Subjects: multi-object tracking -- Kalman filtering -- spatiograms -- occlusion prediction; Evolutionary computation -- Periodicals; Neural networks (Computer science) -- Periodicals; Genetic programming (Computer science) -- Periodicals; Biologically-inspired computing -- Periodicals; Swarm intelligence -- Periodicals; Quantum computers -- Periodicals; 006.3
• Languages: English
• ISSNs: 1751-648X

Abstract

Aiming at the tracking failure caused by occlusion between objects, interleaving or target drift in multi-object tracking process, and the new improved algorithm of occlusion prediction tracking based on Kalman filtering and spatiograms was proposed. By combining the colour histogram and the distribution of colour in space, spatiograms can be used to distinguish between different objects so that we can track the object when interleaving or occlusion between objects occurs. The state of the object can be predicted by the Kalman filtering, and the occlusion mark is used for the object which overlaps with other objects, so that the occluded object which is undetected can be tracked in the next frame video. The 2D MOT 2015 dataset was used in the experimental procedure, and the average accuracy of tracking was 34.1%. The experimental results have shown that the proposed algorithm can improve the performance of multi-object tracking process.