JP3841246B2 - Video hierarchical encoding device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a video layer encoding apparatus, and more particularly to a video layer encoding apparatus suitable for use in digital compression that enables reproduction of a video signal from a portion of compressed data of a video signal.
[0002]
[Prior art]
Conventionally, as a technique for hierarchically encoding a video signal into a base layer and an enhancement layer, there is a scalability technique defined by MPEG-2. For example, the scalability technique can reproduce an image with a low spatial resolution when a part of an encoded bit stream is decoded, and further improve the spatial resolution by decoding the remaining data in the bit stream. This is a technique for obtaining an additional signal.
[0003]
[Problems to be solved by the invention]
The scalability technique defined in the above MPEG-2 is based on the premise that all frames of a moving image are encoded, and there is no frame skipping (frame skip) required in low bit rate video transmission. There is no consideration. In addition, the scalability technique has a problem in that it is relatively difficult to process software on a personal computer because of complicated computation.
[0004]
An object of the present invention is to provide a video hierarchical encoding device using frame thinning that can efficiently transmit image information. Another object of the present invention is to provide a video layer encoding apparatus suitable for video transmission over a low-speed line such as the Internet.
[0005]
To achieve the above object, according to the present invention, in a video layer encoding apparatus that selectively encodes an image frame, first layer encoding means for encoding a first layer, the first layer encoding means, A second layer encoding unit that performs encoding of a layer higher than the layer, and an image frame encoded by the second layer encoding unit is overlapped with the image frame encoded by the first layer encoding unit. Frame selection means for selecting so as not to become, the frame selection means from the frame rate after adding the frame rate of the second hierarchy encoding means to the frame rate of the first hierarchy encoding means , calculated frame interval by using the input image frame rate, so as to select the encoding target frame of the second hierarchical encoding means based on the frame interval obtained, already at the first hierarchical encoding means Starting from the encoded frame, it is characterized in that the frame apart the frame interval from the frame in that set as the encoding target frame of the second hierarchical encoding means.
[0007]
According to the above feature, frames in the upper or second layer do not overlap with frames in the lower or first layer, so that the frame rate to be transmitted is increased, and image quality is improved when these frames are decoded. Will be able to.
[0008]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, the present invention will be described in detail with reference to the drawings. First, the principle of the first embodiment of the present invention will be described with reference to FIG.
[0009]
This embodiment is based on H.264 as the basic hierarchy (BL). 263 baseline coding is performed, SNR scalability is used as the first enhancement layer (EL1), and bi-directional predictive coding (B-picture) is performed as the second enhancement layer (EL2). Here, SNR scalability predicts and encodes information for recovering information lost due to encoding of the base layer. In the figure, the symbol “→” means predictive coding. The same applies to the following figures.
[0010]
According to this embodiment, it is possible to improve the frame image quality by adding the first extension layer to the basic layer. Also, since the frames in the second enhancement layer are selected (B-picture) by selecting frames that were not selected in the base layer and the first enhancement layer, when the second enhancement layer is further decoded. The frame rate is increased and the image quality can be improved. The method of selecting the second enhancement layer frame will be described later with reference to FIG.
[0011]
Next, the principle of the second embodiment of the present invention will be described with reference to FIG. In this embodiment, the encoding of the basic layer and the first enhancement layer is the same as that of the first embodiment, but the second enhancement layer is independent of these layers and the frames not selected in these layers. This is characterized in that predictive encoding is performed by selecting. Prediction coding of the second enhancement layer is performed by intra coding (I-picture) and forward predictive coding (P-picture). Also in this embodiment, when the second enhancement layer is further decoded, the frame rate increases and the image quality can be improved.
[0012]
Next, the principle of the third embodiment of the present invention will be described with reference to FIG. This embodiment is characterized in that each of the base layer, the first enhancement layer, and the second enhancement layer is independent, and the frames that are not selected in the lower layers are selected and predictively encoded. There is. According to this embodiment, when the first enhancement layer and the second enhancement layer are decoded, the frame rate can be improved to two times and three times that when only the basic layer is decoded.
[0013]
Next, a specific example of means for determining the encoding target frame in each embodiment will be described with reference to FIG. The solid line in the figure represents the flow of data, and the dotted line represents the flow of control signals and parameters.
[0014]
The image data to be encoded is input to the base layer encoding device 11 and the enhancement layer encoding device 12. Examples of the base layer encoding apparatus 11 include H.264. H.263 baseline coding can be used. The base layer coding apparatus 11 has a frame skip control mechanism therein, and performs frame skip control in accordance with, for example, an algorithm defined in TMN, which is a publicly known technique published by the ITU. Data encoded by the base layer encoding apparatus 11 is sent out on the line 13. In addition, the base layer encoding apparatus 11 outputs the frame rate (frames / second) in the base layer and the encoded frame information a to the encoding target frame determining apparatus 14.
[0015]
Examples of the enhancement layer encoding device 12 include H.264. Bi-directional encoding, which is temporal scalability defined in H.263, H.263 baseline coding or the like can be used. The enhancement layer encoding apparatus 12 encodes an image frame in accordance with an instruction from the encoding target frame determination apparatus 14. Further, when performing this encoding, the base layer encoding apparatus 11 receives a notification of the image quality parameter b, for example, the average quantization step size, etc., and performs encoding so that the image quality is uniform between the base layer and the extension layer. To do. Then, the frame rate c is calculated from the amount of information e generated as a result of encoding and a predetermined target bit rate f, and notified to the encoding target frame determination device 14. The frame rate c can be obtained by c = f / e, for example.
[0016]
When receiving the frame rate of, for example, A frame / second from the base layer encoding device 11 and B frame / second from the enhancement layer encoding device 12, the encoding target frame determination device 14 receives the average frame when both layers are combined. Calculate rate (A + B) frames / second and average frame interval 30 / (A + B). Here, 30 is the number of frames per second. Then, starting from the frame encoded in the base layer or the frame encoded in the enhancement layer, a frame separated by 30 / (A + B) frames is determined as a frame to be encoded in the enhancement layer. At this time, if 30 / (A + B) is not an integer, it may be used by rounding to the nearest integer. Further, when the selected frame is closer to the frame encoded in the base layer, for example, {30 / (A + B)} / n frame, the frame is excluded from the encoding target. Here, n may be a real number larger than 1, but 2 is preferable, for example.
[0017]
A specific example will be described with reference to FIG. 5. Assuming that the frame rate A = 3 in the base layer coding apparatus 11 and the frame rate B = 7 calculated in the enhancement layer coding apparatus 12, 30 / (A + B) is 3, the encoding target frame determination device 14 starts the frame encoded in the basic layer or the frame encoded in the enhancement layer as indicated by the solid line upward arrow in the enhancement layer encoding device 12. As shown, every 3 frames indicate a frame to be encoded. In addition, when the frame to be encoded every 3 frames is designated in this way, the 9th frame, the 19th frame, etc. (indicated by dotted lines) become the corresponding frames, but these frames are encoded in the base layer. Since it is only one frame away from the encoded frame, it is removed from the encoding target frame.
[0018]
As described above, when the frame to be encoded is determined by the enhancement layer encoding apparatus 12 using the embodiment of FIG. 4, the selected frame is not biased while keeping the frame rate of the base layer + the enhancement layer. Therefore, when these frames are decoded, better image quality than before can be obtained.
[0019]
【The invention's effect】
As is clear from the above description, according to the present invention, the upper or second layer encoding target frame can be made different from the lower or first layer encoding target frame, so that it is transmitted. The frame rate can be increased and the image quality can be improved compared to the conventional method.
[0020]
Also, starting from the encoded frame of the lower or first layer, the frame separated by the average frame interval is set as the encoding target frame, so that the encoding target frame of the upper or second layer is changed to the entire frame. The image frame can be selected or extracted in a balanced manner. In addition, operations are greatly simplified compared to conventional scalability techniques. Further, when the present invention is applied to video transmission on a low-speed line such as the Internet, the effect is great.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram of a first embodiment of the present invention.
FIG. 2 is an explanatory diagram of a second embodiment of the present invention.
FIG. 3 is an explanatory diagram of a third embodiment of the present invention.
FIG. 4 is a block diagram showing a specific example of encoding target frame determination means.
FIG. 5 is an explanatory diagram of an encoding target frame determined by the encoding target frame determination unit.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 11 ... Base layer encoding apparatus, 12 ... Extension layer encoding apparatus, 14 ... Encoding object frame determination apparatus.

Claims (2)

In a video hierarchical encoding device that selectively encodes image frames,
First layer encoding means for encoding the first layer;
Second layer encoding means for encoding a layer higher than the first layer;
Frame selecting means for selecting an image frame encoded by the second hierarchical encoding means so as not to overlap with an image frame encoded by the first hierarchical encoding means;
The frame selection means obtains a frame interval using an input image frame rate from a frame rate obtained by adding the frame rate of the second hierarchy encoding means to the frame rate of the first hierarchy encoding means, Based on the obtained frame interval, the encoding target frame of the second layer encoding unit is selected, and the frame already encoded by the first layer encoding unit is used as a starting point, and only the frame interval from the frame is determined. A video layer encoding apparatus characterized in that a distant frame is set as a frame to be encoded by the second layer encoding means. The video hierarchical encoding device according to claim 1, wherein
When the frame interval between the encoding target frame of the first layer encoding unit and the target frame to be encoded by the second layer encoding unit is smaller than the frame interval, the frame obtained at the frame interval is A video layer encoding apparatus characterized in that the second layer encoding means does not set the frame to be encoded.

Images