Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1998-03-02
2001-02-20
Lee, Richard (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240240
Reexamination Certificate
active
06192078
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a motion picture converting apparatus for a motion picture compression format conversion. More specifically, it relates to a motion picture converting apparatus which is intended to improve a speed of a motion picture format conversion by reducing an amount of operation for a motion vector detection.
2. Description of the Prior Art
Recently, a digitization of a telecommunication technique has been advanced, and a plan is also in progress to digitize a current analog broadcasting system such as NTSC/PAL/SECAM. A new broadcasting service, which a digital technique is applied to, is a direct broadcasting by satellite for providing a multi-channel service, a ground wave broadcasting for a high definition television broadcasting (ATV: Advanced Television), a video on demand or the like. However, when such a motion picture is communicated by a digital signal, a common problem is a large amount of information. A motion picture image compression technique is essential in order to perform a real-time reproduction at a high frame rate and to suppress a traffic of a network.
In order to efficiently compress motion picture image data having a large amount of information, an approach for reducing redundancy by the use of a time-axis direction or a correlation in the time-axis direction is used. A typical approach is an MPEG (Motion Picture Image Coding Experts Group) system. This is discussed by ISO-IEC/JTC1/SC2/WG11 and is proposed as standards. In this system, adopted is a hybrid system of a combination of a motion compensating coding, a DCT (Discrete Cosine Transform) coding and a variable length coding (VLC). A constitution of a typical MPEG coding apparatus is shown in
FIG. 15. A
current frame is input to a motion vector detecting section and a subtracter. In the motion vector detecting section, the current frame is divided into rectangular blocks. By calculating the correlation between each block and a previous frame previously held in a frame memory, a motion vector is detected for each block. A predicted image generated by shifting the previous frame from the location of the current frame block to the point which the motion vector indicates is then input to the subtracter. In the subtracter, a differential image between the current frame and the predicted image is calculated. The differential image is then discrete-cosine transformed and quantized by DCT means and quantizing means respectively. The data is variable-length coded by variable length coding means and is then output outward. At this time, the motion vector detected by the motion vector detecting section is also coded by motion information coding means and is then output outward together with the data. At the same time that the quantized data is input to the variable length coding means, it is also input to inverse quantizing means. The data is restored into the differential image via inverse DCT means. In an adder, the current frame predicted image and the differential image are added to each other, so that the current frame is reconstituted. The reconstituted frame is stored in the frame memory in preparation for the subsequent frame coding.
In such a series of coding processing, it takes much of time for the motion vector detecting section to perform an operation. A study has been heretofore undertaken to reduce or simplify this processing. However, since this reduction or simplification is adapted in an equivalent manner to all macro blocks, a motion vector detection precision is deteriorated. The motion vector is defined so that the processing unit thereof may be set to the macro block (16×16 pixels). The macro block is classified into three types (Intra: the block itself is used because of a large motion; Non-MC coded: a simple difference between frames is operated because of less motion; and MC coded: a normal motion compensation) in accordance with the motion vector thereof. Known is a method disclosed in Japanese Patent Application Laid-open No. 4-181888 in which a block determining circuit is provided before the motion vector detecting section whereby the macro block is classified so that the processing is changed in response to a classifying result. However, when the motion is determined by this method, the block is compared to the block alone positioned in the macro block itself (corresponding to the motion vector 0) in the previous frame. The blocks are simply classified by whether or not there is motion. When a block is determination as having no motion, the motion detection is not performed for this block.
On the other hand, a digital VTR is one of important factors of a digital broadcasting communication technique as described above. An uncompressed digital VTR has been heretofore used as the digital VTR for a broadcasting station in the broadcasting station or the like. The digital VTR has an advantage in that it has a high quality and a deterioration is not noticeable during a dubbing. However, the digitization causes a problem in which recorded information is increased. Moreover, since the digital VTR is large-sized and expensive, it is less generally used. However, recently, advance in a band compression technique and improvement of recording density allow a small-sized/household digital VTR to be developed. A DVC (Digital Video Cassette) format standardized by HD digital VTR conference is used for this digital VTR. DVC data is characterized by the following manner. That is, the data is compressed in a spatial direction by the use of DCT, the data is then variable-length coded by Huffman coding. Although the DVC data has a large amount of information because of compression without the use of a time-direction correlation, it has the high quality and it is easily edited for each frame. Furthermore, since the DVC data is adapted so that it may be captured into a computer, it is expected that the DVC data is more frequently used in the future. In the system such as the video on demand for exchanging the motion picture image through the network, it is difficult to transmit/receive the DVC data at real time depending on a transmission performance of the network. In this case, the method, in which the motion picture image is stored as the DVC data and the data is converted into an image format such as MPEG when required, is effective in view of an effective use of a disk resource. Although a motion picture format conversion has been heretofore performed by the use of an exclusive hardware, it is desired that the motion picture format conversion is performed by the use of a software in consideration of an image quality, a cost, a flexibility toward various formats, a labor of implementing into a computer, a compatibility with the computer, an ease-to-distribute or the like.
However, there is a problem in which it takes a long processing time for the software to decode the DVC data and encode into MPEG at one time. In view of the above problem, it is an object of the present invention to reduce a load applied to a motion vector detection in MPEG encode and to improve a conversion speed of a motion picture conversion. As the prior-art motion vector detection processing, known is a method disclosed in Japanese Patent Application Laid-open No. 4-181888 in which a block determining circuit is provided prior to the motion vector detection in the MPEG encode whereby a macro block is classified so that the processing is changed in response to a classifying result. However, this method involves the following problems:
1. A classifying by a block determining circuit is of two types alone (non-MC coded and others). Although macro blocks are classified into three types of blocks in MPEG, classification is not performed in connection with Intra where motion vector is not found in previous frames.
2. A determination of the block determining circuit as to whether or not the motion is detected is made in accordance with a simple difference between blocks, which causes the problem about detection precision of a motion vector whose
Komiya Daisaku
Uenoyama Tsutomu
Lee Richard
Matsushita Electric - Industrial Co., Ltd.
Pollock Vande Sande & Amernick
LandOfFree
Motion picture converting apparatus 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 picture converting apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Motion picture converting apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2559026