基于均值漂移的目标跟踪算法分析-机械电子工程专业论文.docx

AbstractIn recent years, the intelligent video surveillance system has become more and more important in daily life, and it is also an important research direction in computer vision. Its main task is to detect, identify and track the targets of interest from the video stream, and analyze, understand their behaviors. Moving target tracking plays a key role in the intelligent video surveillance system, and it is also an important indicator to measure the quality of a surveillance system. So it has important research value.Mean Shift is a kernel function based non-parametric estimation algorithm. For the advantages of no prior knowledge requirement, fast convergence, and real time tracking ability, it has attracked lots of interests from researchers. However, there are still many drawbacks to be concerned before it can be put into practice. Therefore, this thises concentrated on the study of Mean Shift so as to improve its target tracking performance. The innovations and main contributions of this thesis are as follows:The multi-feature fusion Mean Shift tracking algorithm is studied, which represents a target through LTP texture and color information to increase robustness. Since LTP sets the noise threshold to a fixed value, it is sensitive to noise variation. To deal with this problem, the LMedS algorithm is introduced for adaptive threshold eatimation to calculate the LTP texture features. Furthermore, tracking window size is adjusted to tightly enclose the target. Experimental results show the effectiveness and robustness of the algorithm.An improved background-weighted Mean Shift tracking algorithm which can effectively reduce background’s interference in target localization is proposed. Due to the constant or the threshold decision changing background histogram can’t effectively represents real time changing background information, so the real-time update background histogram is proposed, while the Kalman filtering algorithm is used for target pre-position. Exp