Method and apparatus for arranging pixels of picture in storage units each having storage size not divisible by pixel size

1. A buffer write method for a buffer including a plurality of M-bit storage units, comprising:

obtaining pixel data of a plurality of first N-bit pixels of a picture;

calculating a corresponding start address of the buffer for the pixel data of the first N-bit pixels; and

storing the first N-bit pixels of the picture according to the calculated start address of the buffer in the M-bit storage units by a buffer controller;

wherein the storing step comprises fully storing at least one of the first N-bit pixels in one of the M-bit storage units storage units; and wherein M and N are positive integers, and M is not divisible by N.

2. The buffer write method of claim 1 , further comprising:

judging if the first N-bit pixels is across a boundary of the M-bit storage units storage units; and

fully storing the first N-bit pixels by the buffer controller in the M-bit storage units storage units if the first N-bit pixels is not across a boundary of the M-bit storage units storage units.

3. The buffer write method of claim 2 , further comprising:

breaking the first N-bit pixels into a first part and second part of the first N-bit pixels and storing the first and second parts of the first N-bit pixels into two storage units of the M-bit storage units respectively by the buffer controller if the first N-bit pixels is across a boundary of the M-bit storage units storage units.

4. The buffer write method of claim 3 , wherein If the picture width of the picture is PW pixel and PW*N is not fully divisible by M, the method further comprises:

judging if a current pixel of the first N-bit pixels is the rightmost pixel in a current pixel row of the picture.

5. The buffer write method of claim 4 , the method further comprises:

storing a next pixel data of the first N-bit pixel in another M-bit storage unit of the buffer if the current pixel of the first N-bit pixels is the rightmost pixel in the current pixel row of the picture.

6. The buffer write method of claim 4 , the method further comprises:

storing at least a portion of a next pixel data of the first N-bit pixel to an end of the current pixel in a current M-bit storage unit if the current pixel of the first N-bit pixels is not the rightmost pixel in the current pixel row of the picture.

7. The buffer write method of claim 1 , wherein the storing step further comprising:

storing P pixels to be fully stored in a first storage unit of the M-bit storage units storage units, where P is a positive integer equal to int (M/N), and int( ) is an integer division with truncation of division result toward zero.

8. The buffer write method of claim 7 , further comprising:

inserting (M−N*P) don't care bits at unused bit locations in the first storage unit of the M-bit storage units storage units.

9. The buffer write method of claim 1 , further comprising:

obtaining pixel data of a plurality of second N-bit pixels of the picture,

wherein the storing step comprises fully storing at least one of the second N-bit pixels in one of the M-bit storage units; the first N-bit pixels are luma pixels, and the second N-bit pixels are chroma pixels.

10. The buffer write method of claim 1 , wherein the storing step is performed based on a raster-scan order of the picture.

11. A buffer read method for a buffer including a plurality of M-bit storage units by a buffer controller, comprising:

calculating a corresponding start address of the buffer for retrieving pixel data of first N-bit pixels of a picture;

calculating a size of the pixel data of the first N-bit pixels;

determining a plurality of addresses of the of M-bit storage units for the first N-bit pixels; and

fetching data storing the M-bit storage units for the first N-bit pixels according to the determined addresses of the of M-bit storage units as the first N-bit pixels;

wherein M and N are positive integers, M is not divisible by N, and each of the M-bit storage units of the buffer stores an integer number of first N-bit pixels without any fractional first N-bit pixel.

12. The buffer read method of claim 11 , wherein the storing step is performed based on a raster-scan order of the picture.