Chinese Journal of Liquid Crystals and Displays, Volume. 38, Issue 8, 1118(2023)
Target tracking method based on improved ECO-HC
Gang-yang LIU1,2、*, Bo HU3, and Yu-qing WANG1
Author Affiliations
1Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China2University of the Chinese Academy of Sciences, Beijing 100049, China3The 32035th Unit of the Chinese People’s Liberation Army, Xi'an 710000, Chinashow less
The tracking method based on correlation filtering has fast frame rate but is prone to tracking drift, resulting in target loss. Aimming to this problem, taking the correlation filtering method ECO-HC as the baseline, this paper proposes parameter S to describe the change amplitude of the image scene. According to real-time calculation of the value for this parameter and updating the model with different intervals, the adaptive updating model is realized, so that the algorithm runs at a higher frame rate and has a good accuracy. This effect is more obvious in the case of complex image illumination changing, target deformation or occlusion. Experimental results show that success rate and accuracy rate of ECO-HC embedded in this method are improved by 1.6% and 2.2% under the condition of changing illumination. When the target leaves the field of view, the success rate is improved by 3.7%, and the accuracy is improved by 3.4%.In this experiment environment, the average frame rate reaches 60 FPS.
