Title :
Object tracking based on integrating the Genetic algorithm with complex method
Author :
Xiaokun Dong ; Jianzhong Cao ; Hongtao Yang ; Zhi Yu ; Huinan Guo ; Chen Liu
Author_Institution :
Xi´an Inst. of Opt. & Precision Mech., Xi´an, China
Abstract :
Traditional Genetic algorithm in target tracking process,is quite effective in the global scope to search target position.However, once the background color and light intensity change.To monitor the state of the target depth movement is often difficult to locate the target accurately. However, complex method can make up for the lack of Genetic algorithms. In view of those, this paper integrates Complex method with traditional Genetic algorithms, uses the center coordinates of the candidate target and scaling coefficient to adjust the focus size as a parameter encoding, and takes the similarity between the target area and the target template as a fitness function,then searches the object in the overall and local scope in turn.The experiments show that the fusion algorithm can achieve accurate tracking of the position and the size of the target.
Keywords :
genetic algorithms; object tracking; target tracking; Complex method; candidate target; genetic algorithm; light intensity change; object tracking; parameter encoding; scaling coefficient; target depth movement; target tracking process; Equations; Genetic algorithms; Histograms; Mathematical model; Sociology; Target tracking; Complex method; Edge Orientation Histogram; Template matchingGenetic algorithms;
Conference_Titel :
Intelligent Control and Information Processing (ICICIP), 2013 Fourth International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4673-6248-1
DOI :
10.1109/ICICIP.2013.6568068