Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1994-11-29
2001-04-24
Rao, Andy (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240170
Reexamination Certificate
active
06222882
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a motion estimating method in motion image coding, and more particularly, to a method for performing a correct motion estimation by preventing the adverse effects due to variations in lighting conditions.
A motion image coding system has actually been developed from a differential pulse code modulation (DPCM) system to a motion compensated interframe coding (MCIC) system, and has recently been extended to a model based coding (MBC) system or an object oriented coding (OOC) system.
While only a predicted error is transmitted in the DPCM system, motion information with a predicted error is transmitted in the MCIC system. In the case of MBC or OOC, motion information and shape information with a predicted error can be transmitted.
The coding by the MCIC system is generally performed in units of blocks having a predetermined size (for example, the size of one block is 4×4 pixels). A motion vector of each block and an error signal between a target image and a reconstructed image obtained by motion compensation, are transmitted.
In coding by the OOC system, a background is separated from an object and the motion of the object is analyzed so as to produce parameters. Here, the background represents the region where no motion exists between adjacent frames, and the object represents the region where a motion exists between adjacent frames.
In a conventional motion estimator, a full search method, a three-step search method and other methods have been used for generating a motion vector. In full searching, a block which is most similar to a block of the current image, is searched in a predetermined search window of the previous image, and then the distance between the current block and the searched block is determined as a motion vector (x,y).
The three-step search has the advantage that quick searching can be performed, but has the disadvantage that an inaccurate motion vector is produced compared with the full search method. Therefore, the full search method is used much more often.
However, an accurate estimation by these methods is difficult to obtain if a luminance signal is changed even by a simple two-dimensional motion of an object. Also, if a light source is changed with no motion, an inaccurate motion vector may be produced. The generation of this inaccurate motion vector results in a large amount of error data in an encoder according to the MCIC system, thereby increasing the quantity of data to be encoded. In addition, in an encoder according to the MBC system, the number of groups is increased and the number of intra-blocks is increased, thereby increasing the quantity of data to be encoded.
SUMMARY OF THE INVENTION
Accordingly, the object of the present invention is to provide a motion estimating method for estimating a motion vector correctly, regardless of a change of a luminance signal.
To attain the above object, there is provided a method for estimating a motion according to the present invention by dividing an image frame represented by luminance signals into a plurality of predetermined size blocks, detecting the most similar block in a previous frame with a block in a current frame and producing a corresponding movement distance as a motion vector, comprising the steps of:
producing a mean value of the luminance signal with regard to a block of the current frame;
producing a mean value of the luminance signal with regard to a block to which a motion estimation of the previous frame is applied;
compensating the luminance signal of the block to which the motion estimation of the previous frame is applied such that the two mean values have the same level; and
selecting a block having the minimum sum of differences between luminance signals of each pixel, with regard to the block of the current frame and the block to which a motion estimation of the previous frame is applied, respectively, and producing the moving distance between these blocks as a motion vector.
REFERENCES:
patent: 5157732 (1992-10-01), Ishii
patent: 5162907 (1992-11-01), Keating et al.
patent: 5361105 (1994-11-01), Iu
patent: 6097832 (2000-08-01), Guillotel et al.
patent: 4-196794 (1992-07-01), None
patent: 93-703651 (1993-11-01), None
Lee Shi-hwa
Shin Jae-seob
Leydig Voit & Mayer Ltd
Rao Andy
Samsung Electronics Co,. Ltd.
LandOfFree
Adaptive motion estimation method of a luminance signal does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Adaptive motion estimation method of a luminance signal, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adaptive motion estimation method of a luminance signal will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2503746