Mutaz. S. AbdulWahab
Abstract
In this paper, a new algorithm for accurate optical flow estimation using discretewavelet approximation is proposed. The image sequences are always assumed tobe noiseless in the computation ...
Read More ...
In this paper, a new algorithm for accurate optical flow estimation using discretewavelet approximation is proposed. The image sequences are always assumed tobe noiseless in the computation of optical flow, since there is always a method thatcan perform such task. One of the main application areas of the wavelettransform is that of noise reduction in images. The basic technique is to transformthe noisy input image into a domain, in which the main signal energy isconcentrated into as few coefficients as possible, while the noise energy isdistributed more uniformly over all coefficients. The choice of the transform isrepresent an important tool in optical flow estimation. In this paper, severalalgorithms of 1-D, 2-D and 3-D wavelet transforms are adapted for the estimationof optical flow for the first time.