Abstract: Since the single feature target tracking algorithm cannot adapt to the change of complex scene well,it is easily affected by the influence of scale variation,deformation,occlusion and background mixed and so on,which leads to the failure of tracking.This paper proposes a multi-scale correlation filter tracking algorithm based on adaptive feature fusion.Firstly,according to the HOG and CN features of the target,the context-aware correlation filter framework was adopted to get the filtering response values of the two features,and the two response values were subsequently normalized.Secondly,according to the distribution weight of the response value and the linear weighted fusion,the fusion response value was calculated to determine the target location.At the same time,scale correlation filter was introduced to estimate target scale changes and enhance scale adaptability capacity.Finally,the quality of model updating was improved by using the predefined response threshold as the judgment condition of translation and scale filter model update.The OTB Benchmark data set is used to test the proposed algorithm,and the experiments are compared with the 11 target tracking algorithms based on the correlation filtering and the context aware framework.The experimental results show that the proposed algorithm achieves promising tracking results,where the distance precision rate is 82.5%and the overlap success rate is 54.2%,and has strong robustness in sophisticated scenarios such as scale variation,deformation,fast motion,occlusion,and so on.