Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2000-03-08
2003-04-22
Philippe, Gims S. (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C348S398100, C382S240000, C382S281000
Reexamination Certificate
active
06553071
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an apparatus capable of processing a motion picture using a wavelet transformation and a method thereof, and in particular to a motion compensation coding apparatus using a wavelet transformation and a method thereof which are capable of detecting a motion vector with respect to a block having a certain change or a motion in an image from a region having a hierarchical structure based on each frequency band and each sub-frequency band generated by wavelet-transforming an inputted motion picture and effectively coding a motion using the detected motion vector.
2. Description of the Background Art
Recently, a computer communication network is advanced, a data including much data such as a motion picture is transferred through the above-described computer communication network. In order to effectively transfer the above-described data, an advanced motion picture compression coding method and transfer technique are increasingly needed. For example, the Motion Picture Expert Group (MPEG) is one of the standards of the motion picture compression coding method. A known motion picture compression method will be explained.
A motion picture is formed of frames which are changed based on the elapse of time or is formed of video signal sequences. In order to effectively compress the motion picture, a two-dimensional space overlap of an image and a time axis-base overlap must be eliminated.
The two-dimensional space overlap of the image data may be decreased by processing a difference image corresponding to a difference between images based on a DCT (Discrete cosine Transform), and the overlap of the time axis may be decreased by compensating a motion with respect to a block having a motion at a center portion having a largest motion of an object among continuous images with respect to the time axis. Two neighboring images are compared with respect to time, and an image data of a previous image is used for processing the current image with respect to the portion in which there is no change in the image or the portion in which there is a small change in the image, and the portion having a large changes is motion-compensated for thereby significantly decreasing the amount of the image data which will be compressed and transmitted.
Here, a process for searching similar blocks in the previous images and the current images is called as a motion estimation by comparing the images because a change is small between the above-described images. In addition, a displacement of the motion in the blocks is called as a motion vector.
In the image compression method used for the MPEG, the images are compressed by the motion compensated DCT which compensates the motions of the block using only an image signal and a luminance signal in order to eliminate the overlap of the time and space. Namely, in the method for coding the images based on the MPEG, there are an interceding method and an intracoding method. The interceding method is directed to compensating the motion with respect to the inputted image and recovering the image and then obtaining a difference image by subtracting the recovered image from the original image for thereby performing DCT with respect to the difference image and performing the coding operation. In addition, the intracoding is directed to performing DCT and the coding operation without motion-compensating the image having large change and the macro block.
In the method for detecting a change of the image and recovering to the original image with respect to the motion picture compressed by MPEG or Motion JPEG based on the interceding and intracoding operations, there are a method for directly processing the compressed images and a method for fully recovering the compressed images. The method for directly processing the compressed images is capable of decreasing an operation time but the accuracy is decreased. The method for fully recovering the images is capable of accurately detecting the images but the processing speed is slow. Namely, since all image data are searched by comparing the original images having a space overlap by the block unit, the process time is increased.
Recently, in order to overcome the above-described problems and a blocking effect which is a disadvantage of the DCT and a resolution degradation which occurs when increasing the compression rate of the images, the images are compressed and coded using the wavelet based transformation.
FIG. 1
is a view illustrating the construction of a Discrete Wavelet Transformation for processing the image signals in the conventional art which includes a wavelet transform coding step(Discrete Wavelet Transformation) for down-sampling the inputted image signal and analyzing into wavelet coefficients, and an inverse wavelet transformation coding step(Inverse Discrete Wavelet Transformation) for up-sampling the down sampled wavelet coefficients and recovering to the original image signal.
The wavelet transformation coding step down-samples the inputted image signals to the signal which is one half of the previous signal using a low frequency analyzing filter H
0
and a high frequency analyzing filter H
1
and a down-samples the signals to the signals which is one half of the previous signal using the low frequency signal H
0
and the high frequency analyzing filter H
1
for thereby generating a wavelet transformation coefficient and analyzing the inputted image signal to the down sampled wavelet coefficients.
The inverse wavelet transformation coding step(Inverse Discrete Wavelet Transformation) up-samples the down sampled wavelet transformation coefficients to the signals which is twice the previous signal using a low frequency combining filter G
0
and a high frequency combining filter G
1
for thereby generating an up-sampled wavelet transformation coefficient and repeatedly performing the up-sampling processes, so that the images are combined to the original images.
Here, since the outputted image signals are processed based, on the analyzing filters H
0
, H
1
, G
0
and G
1
, the input image signals are delayed and outputted later compared to when the inputted image signals are not processed by the analyzing filters and outputted. However, there is no difference in the contents of the images contained in the input images and output images.
The wavelet transformation method is directed to independently dividing the inputted signals by the band, compressing and coding the same. During the coding operation, the sub-band coding itself hierarchically divides the data into different resolutions and adds and subtracts the signals of the added hierarchies between nodes having different resolutions for thereby using a desired band width.
The inputted video signals is divided into different frequency bands by the band analyzing filter and is divided into frequency bands of two-dimensional space. As the frequency band is increased, the importance of the divided band is decreased during the reproducing operation. Actually, in the case of the image data of the low frequency and, an image having a smoothed boundary of an object is outputted, and in the case of the high frequency band having a high frequency, only vertical boundary, horizontal boundary and diagonal components are outputted based on the frequency band.
FIG. 2
is a view illustrating the distribution of a sub-frequency band of a wavelet coefficient which is wavelet-transformed and then generated. The sub-frequency band HLi positioned at the right upper portion represents a high frequency component of the horizontal direction, and the sub-frequency band LHi positioned at the left lower portion represents a high frequency component of the vertical direction, and the sub-frequency band HHi positioned at the right lower portion represents a high frequency component of the diagonal direction. In particular, since the visual ability of a human is weak with respect to the diagonal components, the sub-band region of the diagonal band may be more in detail divided. Namely, the sub-frequency reg
Kim Jeong Min
Park Young Man
Fleshner & Kim LLP
LG Electronics Inc.
Philippe Gims S.
LandOfFree
Motion compensation coding apparatus using wavelet... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Motion compensation coding apparatus using wavelet..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Motion compensation coding apparatus using wavelet... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3052606