M section 论文写作.pptxVIP

M section3D Motion Estimation for Depth Image Coding in 3D Video Coding;Structure—— A. 3D block matching algorithm; The value of x and y are indicated by either integer or non-integer number of pixels, which is called sub-pixel precision.⑤ z component represents the shifting direction in depth dimension that corresponds to the pixel value.⑥ In case of 8 bits per pixel, the depth value can be varied between 0- 255. ⑦ Therefore a motion vector in z component can be an integer value in the range of -255 to +255. ⑧ Finally, the values of depth after motion compensation are limited by 0 ≤ f(x, y) + z ≤ 255 since the pixel value is represented by an unsigned 8 bits number. ⑨ f(x, y) is a pixel value in the reference frame. ⑩ The 3D search space for depth frames is shown in Fig. 3. 11; To perform 3D-BM, the block matching is firstly performed in two dimensions (Fig. 2.): horizontally and vertically. ① The searching algorithm is performed over the reference frame within the searching window to minimize the motion prediction error in x and y components. ② The matching function of mean squared error is defined as follows: …… where –w i, j w and w is the size of searching window. ③ The f(m, n) represents the current block of M × N pixels at coordinates (m, n) and g(m + i, n + j) represents the corresponding block in the reference frame at new coordinates. ④ At the best match position of i = x and j = y, the motion vector MV(x, y) represents the shift distance in horizontal and vertical directions of all pixels within the block. ⑤; After obtaining the best match in x and y components, the block matching in depth direction is then performed by varying each pixel value in the block between zmin to zmax, where zmin, zmax denote the upper and lower search boundary in the depth dimension. ⑥ Note that new pixel value must be in the range of 0-255. ⑦ Finally, the value between zmin and zmax which minimizes the motion compensation error is select