Moving object segmentation using the 9/7–10/8 dual-tree complex filter bank
Aryaz Baradarani, Q. M. Jonathan Wu · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · 2008
This paper introduces our recently designed 9/7-10/8 dual-tree complex filter bank and studies its application in moving object detection and segmentation. The dual-tree complex wavelet transform (DT-CWT) is used in edge detection required for segmentation in video frames. The DT-CWT approach is compared with the previous methods based on scalar wavelets and multi-wavelets (MW), using double change detection (DCD) technique. The results show that DT-CWT with the 9/7-10/8 filter bank compare favorably to the previous results in terms of the number of detected edges and the correctness of the segmented area.