Combined improved Frequency-Tuned with GMM algorithm for moving target detection
Hui Wang, Jing Peng Gao, Lijun Yu, Yukun Hu, Zhengan Wang · 2017
A new moving target detection algorithm is proposed by the integration of improved Frequency-Tuned and GMM (Gaussian Mixture Model) algorithm. This algorithm aims at the problem of random noise in detecting dynamic significant information by Frequency-Tuned(FT) algorithm and the problem of large computation in detecting the moving target by GMM algorithm. Firstly, FT algorithm was improved from two aspects: distance metric and feature fusion. Then the dynamic saliency map was detected by the improved FT algorithm. Finally, the detected saliency map could be further detected by Gaussian Mixture Model to get the final conclusions. The experimental results prove that the proposed method can avoid the random noise and highlight the significance of the moving target, improve the detection rate with good robustness.