In this paper a novel scheme for video object extraction based on the second generation of the curvelet transform which is called Fast Discrete Curvelet Transform (FDCT) via wrapping is introduced. The main advantages of using FDCT are the approximate properties and the high directional sensitivity of this transform. An imaginary component of the curvelet coefficients to extract the moving objects in the video sequence is used. The proposed algorithm is mainly divided into two steps. The first step is based on estimating the static background from the initial frames using FDCT, the seconded step is based on subtracting the background from each video frame for obtaining the moving objects. Experimental results show a promising results than traditional wavlelet transform, where the accuracy ratios of each video frame is maximized while the error ratios are minimized compared with the wavelet transform.
Digital Object Identifier (DOI)
Sayed, Usama; A. Mofaddel, Mahmoud; and M. Abd-elgawad, Marwa
"Video Object Extraction using the Curvelet Transform,"
Applied Mathematics & Information Sciences: Vol. 10:
2, Article 21.
Available at: https://digitalcommons.aaru.edu.jo/amis/vol10/iss2/21