Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2000-03-09
2003-09-23
Kelley, Chris (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240160, C375S240240, C382S239000, C382S232000, C382S238000, C382S236000, C348S699000, C348S595000
Reexamination Certificate
active
06625214
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to apparatus and method for an information encoding for performing inter-frame prediction on the basis of motion compensation and intra-frame encoding.
The present invention also relates to apparatus and method for a fade detecting in the information encoding apparatus and method.
2. Description of the Related Art
There are various encoding methods proposed for use of encoding information, especially digital information, which is, for example, stored on a recording medium, or transmitted on a broadcast wave. However, MPEG (moving picture coding expert group) standard is used for recording a moving-picture data on a DVD having about seven times capacity as compared with that of a conventional recording medium such as a CD (Compact disk).
The MPEG standard is an encoding system in which DCT(Discrete Cosine Transform), motion compensation prediction and variable length encoding are combined to carry out compression of an enormous amount of picture data efficiently. DCT processing is performed with respect to predictive error data which is determined as a difference between a current objective frame and a forward or a backward reference frame delivered by means of motion compensation. Then the predictive error data encoded by a DCT is quantized. Additionally, motion vector detection is performed. In the MPEG standard, the frame is divided into macroblocks. Each macroblock includes 16 pixels.times.16 pixels. In the case where any motion is detected every macroblock between the reference frame and the current objective frame, a direction and an amount of the motion is indicated as a motion vector. Then a predictive picture is generated on the basis of the motion vector, the difference at a macroblock between the predictive picture and a reference picture, which generates the motion, is calculating. Further, the difference is encoded by variable length encoding. In this manner, an enormous amount of picture data is compressed efficiently.
However, according to a conventional compression method, in the case where both of a monotonous picture and a complicated picture are compressed at same compression ratio, the complicated picture is compressed so excessively that the compressed complicated picture declines. Therefore, a deterioration of picture quality is occurred.
For example, when a picture having a change of a luminance, so called a fade, whose value of luminance is changed in one direction such as from minimum to maximum and vice versa, is compressed at low bit rate on the basis of MPEG standard, It is known that “block distortion” is often generated.
Therefore, in the conventional compression method, a fade deteccting is executed. If the fade is detected, the bit rate is adapted to the fade, or a motion compensation is adapted to the fade.
However, in the conventional fade detection method, the fade detection is executed by means of calculating an accumulation of value of luminance in each field is calculated, and measuring a change of the accumulation in time domain for several fields. Then fade is determined when the accumulation changes gradually.
In the case of measuring a change of the accumulation in time domain, a delay of frame is occurred until fade is detected, so that some memories for storing the delayed frame are required in order to encode the fade after detected.
Additionally, in accordance with this conventional fade detection method, a scene change or a complicated picture might be detected as fade in error because the fade detection is executed on the basis of a change of value of luminance at each field.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide method and apparatus for a fade detecting, and an information encoding, which are capable to detect fade exactly without extra memories for a fade detecting.
The above object of the present invention can be achieved by an apparatus for detecting fade transitions in image data used for image data compression. The apparatus has an inter-frame difference calculation device, a pixel block detection device, an predictive difference calculation device, and a judging device. The inter-frame difference calculation device calculates an absolute value of a difference at each pixel between two frames, and a sum of absolute values of differences at all pixels of one frame. The pixel block detection device calculates an absolute value of a difference at each pixel between two pixel blocks. Each of block is subdivided from one frame. And the pixel block detection device detects a pair of blocks to obtain the minimum value for a sum of absolute values of differences at all pixels of one block. The predictive difference calculation device calculates a total of sums of the absolute values of differences at all pixels of one block in each pair of blocks, which is detected by the pixel block detection device, for one frame. The judging device judges whether the fade transitions are generated or not, on the basis of the sum of the absolute values at all pixels of one frame, which is calculated by the inter-frame difference calculation device, and the total, which is calculated by the predictive difference calculation device.
According to the fade detecting apparatus of the present invention, the inter-frame difference calculation device calculates an absolute value of a difference, such as luminance, chrominance and the like, at each pixel between two frames. Then the inter-frame difference calculation device calculates a sum of absolute values of differences at all pixels of one frame. Therefore, a correlation between the two frames is obtained, a picture having a high correlation is distinguish from a picture having a low correlation. Next, the pixel block detection device calculates an absolute value of a difference at each pixel between two pixel blocks. Each of block is subdivided from one frame, called macroblock. And the pixel block detection device detects a pair of blocks to obtain the minimum value for a sum of absolute values of differences at all pixels of one block. The predictive difference calculation device calculates a total of sums of the absolute values of differences at all pixels of one block in each pair of blocks, which is detected by the pixel block detection device, for one frame. Therefore, the correlation is obtained every macro block, so that, in the case where a picture has some motion, the correlation between the frame is calculated accurately. Next, the judging device judges whether the fade transitions are generated or not, on the basis of the sum of the absolute values at all pixels of one frame, which is calculated by the inter-frame difference calculation device, and the total, which is calculated by the predictive difference calculation device. That is, both a correlation between two frames and a correlation between the two blocks are used for a fade deteccting. First, a fade picture and a normal picture are accurately distinguished from a scene change and a complicated picture, whose correlation between two frames is lower than the normal and the fade pictures, on the basis of a correlation between two frames. Next, a fade picture is accurately distinguished from a normal picture, which has much more motion than the normal picture on the basis of a correlation between the two blocks. Therefore, a fade picture, whose luminance changes toward one direction at a certain rate for a whole frame, is accurately detected without a detection error.
In one aspect of the fade detecting apparatus of the present invention, the judging device has a first compare device and a second compare device. And the judging device judges that the fade transitions are generated in the case where such a status that the sum is smaller than the first reference value as a result of comparing by the first compare device, and the total is greater than the second reference value as a result of comparing by the second compare device, is continued for a plurality of frames.
According to this aspect, the judg
Takahashi Tsutomu
Umehara Yasuyuki
An Shawn S.
Kelley Chris
Pioneer Corporation
Pitney Hardin Kipp & Szuch LLP
LandOfFree
Apparatus and method for fade detecting and information... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for fade detecting and information..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for fade detecting and information... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3101190