Method and apparatus for decoding video bitstreams to reduced spatial resolutions

• US 7,006,572 B2
• Filed: 05/09/2002
• Issued: 02/28/2006
• Est. Priority Date: 05/09/2002
• Status: Expired due to Fees

• Current Assignee: Mitsubishi Electric Research Laboratories, Inc. (Mitsubishi Electric Corporation)
• Original Assignee: Mitsubishi Electric Corporation
• Inventors: Porikli, Fatih M., Sun, Huifang
• Primary Examiner(s): DIEP, NHON THANH
• Application Number:
  US10/141,742
• Publication Number:
  US 20030210744A1
• Time in Patent Office: 1,391 Days
• Field of Search: 375/240.01, 375/240.25, 375/240.24, 375/240.2, 375/240.21, 375/240.16, 348/408.1
• US Class Current: 375/240.16
• CPC Class Codes: H04N 19/122 Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
• CPC Class Codes: H04N 19/129 Scanning of coding units, e.g. zig-zag scan of transform coefficients or flexible macroblock ordering [FMO]
• CPC Class Codes: H04N 19/132 Sampling, masking or truncation of coding units, e.g. adaptive resampling, frame skipping, frame interpolation or high-frequency transform coefficient masking
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/18 the unit being a set of transform coefficients
• CPC Class Codes: H04N 19/44 Decoders specially adapted therefor, e.g. video decoders which are asymmetric with respect to the encoder
• CPC Class Codes: H04N 19/59 involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A method decodes a video encoded as a compressed bitstream including frames constructed as blocks. The bitstream is variable length decoded and inverse quantized to determine motion vectors and DCT coefficients of the blocks of the bitstream. The DCT coefficients are then inverse discrete cosine transformed. A low-resolution block is generated for each reconstructed block, and the reconstructed blocks are motion compensated with the low-resolution blocks to decode the video.

Claims

1. A method for decoding a video encoded as a compressed bitstream encoded as a plurality of blocks, comprising:

variable length decoding and inverse quantizing the bitstream to determine motion vectors and DCT coefficients of the blocks of the bitstream;

inverse discrete cosine transforming the DCT coefficients to produce reconstructed blocks;

generating, for each reconstructed block, a low-resolution block of the DCT coefficients constructed according to an ordered pattern determined by a shape function, wherein a number of DCT coefficients of the low-resolution blocks is less than a number of the DCT coefficients of the reconstructed blocks; and

motion compensating the reconstructed blocks with the low-resolution blocks to decode the video, wherein each reconstructed block includes a plurality of pixels, and the shape function partitions the reconstructed block into covered pixels and uncovered pixels.

2. The method of claim 1 wherein the low-resolution blocks are stored in a frame memory having a constrained size.

3. The method of claim 2 wherein the generating further comprises:

discrete cosine transforming the DCT coefficients according to a constrained size of the frame memory; and

filtering and indexing the transformed DCT coefficients.

4. The method of claim 3 wherein the motion compensating further comprises:

inverse indexing the low-resolution blocks; and

inverse discrete cosine transforming the indexed low-resolution blocks according to the constrained size of the frame memory.

5. The method of claim 1 wherein the shape function ƒ

(u, v) is selected from a group consisting of;

$\begin{array}{c}f<br><br>\left(u,v\right)=\{\begin{array}{cc}1& \frac{u+k}{v+k}=1\\ 0& \mathrm{else},\end{array}\\ f<br><br>\left(u,v\right)=\{\begin{array}{cc}1& \{\begin{array}{c}\left[\left(u=k\right)\bigwedge <br><br>\left(v><br><br>2<br><br>k\right)\right]<br><br><br><br>\mathrm{or}<br><br>\left[\left(v=k\right)\bigwedge <br><br>\left(u><br><br>2<br><br>k\right)\right]<br><br><br><br>\mathrm{or}\\ \left[\left(u=2<br><br>k\right)\bigwedge <br><br>\left(2<br><br>k><br><br>v><br><br>k\right)\right]<br><br><br><br>\mathrm{or}<br><br>\left[\left(v=2<br><br>k\right)\bigwedge <br><br>\left(2<br><br>k><br><br>u><br><br>k\right)\right]\end{array}\\ 0& \mathrm{else},\end{array}\\ f<br><br>\left(u,v\right)=\{\begin{array}{cc}1& u^2+v^2=k^2\\ 0& \mathrm{else},\end{array}\\ f<br><br>\left(u,v\right)=\{\begin{array}{cc}1& \left[\left(u=k\right)\bigwedge <br><br>\left(v<<br><br>k\right)\right]<br><br><br><br>\mathrm{or}<br><br>\left[\left(v=k\right)\bigwedge <br><br>\left(u<<br><br>k\right)\right]\\ 0& \mathrm{else},\end{array}\\ f<br><br>\left(u,v\right)=\{\begin{array}{cc}1& \left[\left(u=k\right)\bigwedge <br><br>\left(v><br><br>k\right)\right]<br><br><br><br>\mathrm{or}<br><br>\left[\left(v=k\right)\bigwedge <br><br>\left(u><br><br>k\right)\right]\\ 0& \mathrm{else},\end{array}<br><br><br><br>\mathrm{and}\\ f<br><br>\left(u,v\right)=\{\begin{array}{cc}1& \begin{array}{c}\left[\left(u=k\right)\bigwedge <br><br>\left(v><br><br>\mathrm{ck}\right)\right]<br><br><br><br>\mathrm{or}<br><br>\left[\left(v=k\right)\bigwedge <br><br>\left(u><br><br>\mathrm{ck}\right)\right]<br><br><br><br>\mathrm{or}\\ u^2+v^2=c^2<br><br>k^2\end{array}\\ 0& \mathrm{else},\end{array}\end{array}$

where k and c are constant numbers, and u and v are coordinates of the DCT coefficients, and dimensions of each block is N×

N.

6. The method of claim 5 wherein the shape function has a center of symmetry such that ƒ

(i, i)=1.

7. The method of claim 5 wherein the ordered pattern of DCT coefficients is

{P(1), P(2), . . . , P(N×

N)}={(u₁, v₁), (u₂, v₂), . . . (u<sub>N×

N</sub>, v<sub>N×

N</sub>)}.

8. The method of claim 7 wherein the first K coefficients are

{P(1), P(2), . . . , P(K)}={(u₁, v₁), (u₂, v₂), . . . (u_K, v_K)}.

9. The method of claim 8 wherein the constrained CDCT for the N ×

N block are $Y<br><br>\left(u_k,v_k\right)=\frac{2<br><br>c_{\mathrm{uk}}<br><br>c_{\mathrm{vk}}}{N}<br><br>\underset{m=0}{\overset{N-1}{\sum <br><br>}}<br><br>\underset{n=0}{\overset{N-1}{\sum <br><br>}}<br><br>y<br><br>\left(m,n\right)<br><br>F_{2<br><br>N}^{\mathrm{uk}}<br><br>\left(m\right)<br><br>F_{2<br><br>N}^{\mathrm{vk}}<br><br>\left(n\right),$

where (u_k, v_k)=P(k), and 1≦

k≦

K.