Image analysis – Image compression or coding – Interframe coding
Reexamination Certificate
1998-05-20
2001-04-17
Couso, Jose L. (Department: 2721)
Image analysis
Image compression or coding
Interframe coding
Reexamination Certificate
active
06219455
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to a method and an apparatus for detecting motion vectors. More particularly, the invention relates to a method and an apparatus for detecting motion vectors in which motion vectors are detected with low power consumption.
2. Description of the Related Art
According to the Moving Picture Experts Group (MPEG) of the International Organization for Standardization/International Electrotechnical Commission, Joint Technical Committee 1/Sub Committee 29) (ISO/IEC JTC1/SC29), moving picture data is compressed by detecting a motion vector. In other words, it is necessary to detect a motion vector in order to perform the above type of compression.
FIG. 15
illustrates an example of a known motion vector detecting apparatus. The configuration of this apparatus is disclosed in pages 61 to 68 of the Technical Report of IEICE ICD93-80 (1993-08) as “half-pel-precision motion-vector detecting LSI in compliance with CCIR601”.
In this example, search window data (pixel data within a search area) is supplied to a calculating unit
71
from an input circuit (not shown). The pixel data is shifted by columns of a reference block from a side register (SR)
81
to a processing element (PE)
82
. Each PE
82
stores pixel data for one search block. Each PE
82
also has predetermined reference block data (template data) for detecting a motion vector. The PE
82
then calculates the difference between the predetermined reference block pixel data and the stored search block pixel data and outputs it to a sum circuit
72
. One stage of cascade-connected PEs
82
sequentially transfers each column of pixel data of a search block (reference block) to the subsequent stage of PEs
82
in synchronization with the clock. Similarly, each stage of PEs
82
transfers pixel data. Then, upon inputting a new column of data, the PE
82
again calculates the difference between the reference block pixel data and the search block pixel data and outputs it to the sum circuit
72
.
The sum circuit
72
obtains the sum of the absolute values of the differences between the reference block pixel data and the search block pixel data in each search block so as to output the sum to a minimum value detecting circuit
73
. The minimum value detecting circuit
73
then detects the position of the search block within the search window obtained when the sum circuit
72
outputs the minimum value.
In this manner, according to the above known type of motion vector detecting apparatus, pixel data is sequentially transferred from one stage of PEs to a subsequent stage of PEs, i.e., pipeline processing is performed to transfer pixel data. Thus, data is overwritten in most of the PEs whenever each row of pixel data is transferred in each clock, causing a current to flow and increasing the power consumption.
SUMMARY OF THE INVENTION
Accordingly, in view of the above background, it is an object of the present invention to suppress the transferring of pixel data within a search area (search data) by interleaving the pixel data in the horizontal direction or in the vertical direction, thereby inhibiting the consumption of a large amount of power.
In order to achieve the above object, according to one aspect of the present invention, there is provided a motion vector detecting apparatus including a first storage unit for storing top-field pixel data of odd-numbered rows or odd-numbered columns among pixel data within a search area. A second storage unit stores top-field pixel data of even-numbered rows or even-numbered columns among pixel data within a search area. A third storage unit stores bottom-field pixel data of odd-numbered rows or odd-numbered columns among pixel data within a search area. A fourth storage unit stores bottom-field pixel data of even-numbered rows or even-numbered columns among pixel data within a search area. A selector selects pixel data of a predetermined search block from the pixel data stored in the first storage unit through the fourth storage unit. A fifth storage unit stores at least interleaved pixel data among pixel data of a reference block. A first calculator performs a calculation to obtain a difference between the pixel data of the search block selected by the selector and the pixel data of the reference block stored in the fifth storage unit. A second calculator performs a calculation to obtain a sum of the differences within each of the search blocks obtained by the first calculator. A detector detects a search block corresponding to a minimum value of the calculations performed by the second calculator.
According to another aspect of the present invention, there is provided a motion vector detecting method including the following steps. In a first storage step, top-field pixel data of odd-numbered rows or odd-numbered columns among pixel data within a search area is stored in a first storage unit. In a second storage step, top-field pixel data of even-numbered rows or even-numbered columns among pixel data within a search area is stored in a second storage unit. In a third storage step, bottom-field pixel data of odd-numbered rows or odd-numbered columns among pixel data within a search area is stored in a third storage unit. In a fourth storage step, bottom-field pixel data of even-numbered rows or even-numbered columns among pixel data within a search area is stored in a fourth storage unit. In a selection step, pixel data of a predetermined search block is selected from the pixel data stored in the first storage unit through the fourth storage unit. In a fifth storage step, at least interleaved pixel data among pixel data of a reference block is stored in a fifth storage unit. In a first calculation step, a calculation is performed to obtain a difference between the pixel data of the selected search block and the pixel data of the reference block stored in the fifth storage unit. In a second calculation step, a calculation is performed to obtain a sum of the differences within each of the search blocks obtained by the first calculation step. In a detection step, a search block corresponding to a minimum value of the calculations performed by the second calculation step is detected.
REFERENCES:
patent: Re. 35093 (1995-11-01), Wang et al.
patent: 4432009 (1984-02-01), Reitmeier et al.
patent: 4603350 (1986-07-01), Arbeiter et al.
patent: 5193004 (1993-03-01), Wang et al.
Couso Jose L.
Frommer William S.
Frommer & Lawrence & Haug LLP
Polito Bruno
Sony Corporation
LandOfFree
Method and apparatus for detecting motion vectors does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for detecting motion vectors, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for detecting motion vectors will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2515222