According to this heuristic, we only consider a limited number of consecutive frames to compute the visible difference prediction map of a specific frame.
Other methods use a sequence of frames to compute properties like speed to draw conclusion about the activity or occurred events [8, 9].
The ACI moment magnifier procedure for slender columns uses the moment magnifier δ ns and the larger of the two column end moments M2 obtained from a conventional elastic frame analysis to compute the magnified moment M c (M max ), which includes second-order effects occurring along the height of the column: M m a x = M c = δ n s M 2 = C m δ 1 M 2 ≥ M 2 (1).
Kurtosis, K [ 24], is defined as where μ4 is the fourth order moment about the mean intensity recorded at pixel (x,y), σ is the standard deviation at pixel (x,y), m is the m-th frame of kurtosis signal, N is the number of frames used to compute the kurtosis, I is the intensity of each frame at pixel (x,y) and μ 0 is the mean between N frames at pixel (x,y).
By comparing the errors obtained for the binding potential, at each voxel, using the framing sequences proposed in [1] and [3], we show that clear differences arise when using different framing sequences to compute kinetic parameters.
The depth information and the respective gradient (computed with depth values of each frame) is used to compute the importance of each frame.
In the frame-based compression, the domain blocks from the previous frame are used to compute the approximate transformation for the range blocks of the current frame.
As usual for WZ video coding, only the luminance component of each frame is used to compute the overall bit rate and PSNR which always considers both the key frames and WZ frames.
The algorithm is described in Algorithm 2. For each frame, we have to compute the eigenvalue decomposition (EVD) and then perform AIC or MDL.
