Target tracking based on non-linear kernel density estimation and Kalman filter
Yang Wu, Xiaofeng Zhou, Yichi Zhang · 2015
This paper chooses Mean Shift algorithm to track target based on non-linear kernel density estimation and Kalman filter. Kernel density estimation is a probability density estimation method, which is used to detect moving target and update the target color histogram. The interest targets are obtained by labeling connected region in the detected binary image. Kalman filtering is employed to predict the position of the target being tracked, giving a starting searching window for Mean Shift tracking. Experimental results show that the method proposed is effective and fast in implementation, which satisfies the real-time requirement, it is capable of handling occlusion problem, meanwhile it is robust against the effects of unstable scene illumination.