JP2000032448A - Image encoded data creation device - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To provide an apparatus for generating encoded image data, which can perform frame thinning for each area depending on video content. SOLUTION: An area selecting section 101 separates encoded data in a frame into encoded data GOB1 in a video area and encoded data GOB2 in a subtitle area. The frame thinning-out changing means 102 discriminates only GOB1 to be thinned or not, and the GOB1 which is not thinned is a GOB merging unit 106.
The GOB 1 to be transmitted to the frame decimation unit is transmitted to the frame decimation unit 104. The frame thinning operation unit 104 thins out GOB1 using the conventional frame thinning process, and sends the frame thinned encoded data to the GOB merge unit 106. The GOB merge unit 106 converts the encoded data GOB2 of the subtitle area into the frame thinning change unit 102
Alternatively, encoded data of one frame is created using the encoded data obtained by the frame thinning operation unit 104.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a technique for creating encoded image data in the field of communication of encoded image data at end-to-end and distribution of encoded image data over a server-client type network.

[0002]

2. Description of the Related Art End-to-end real-time communication of accumulated moving picture encoded data and transmission of moving picture encoded data in an application in which a single video transmitted from a server is shared by a plurality of clients at the same time. When transmitting encoded data via a transmission path having a transmission bit rate different from the encoding bit rate,
In particular, in the case of a transmission path with a small transmission bit rate, in order to reduce the encoded data, frame thinning of the encoded data is performed.

The frame thinning process includes a method of thinning out frames by decoding encoded data and then changing and encoding a reference image, and a method of thinning out frames without decoding encoded data.

[0004] A method of thinning out a frame by once decoding encoded data and changing and encoding a reference image is to decode a frame from a non-thinned frame to a next non-thinned frame and convert the image of the frame to be thinned out. This is a method in which a reference image is set to a frame that is not thinned out, and coding is performed again without using it.

On the other hand, in a method of thinning out frames without decoding encoded data, an algorithm of an inter-frame differential encoding method is used.

In general-purpose coding methods such as MPEG-2, inter-frame difference coding is generally used, and each frame performs inter-frame difference coding according to the time direction of a frame to be referred to. Intra-frame (I-frame) that does not exist, frame (P-frame) that encodes the difference between the previous frame, and frame (B-frame) that encodes the difference between the previous frame and the subsequent frame. Can be classified.

[0007] Of these, the B frame has a property that the receiving side of the encoded data can decode after receiving the data of the frame later than the B frame, and cannot be used as a reference image of another frame. However, the frame thinning process has been performed by simply deleting the B frame data itself. On the other hand, since a P frame can serve as a reference image for a subsequent frame, it is not merely deleted, but a frame that refers to the deleted P frame is coded so that a frame not to be deleted is newly set as a reference image. You need to process the data. For this reason, there is a method of thinning out a frame without going through a decoding process, such as a method of searching for a motion of a DCT coefficient of error information using a property that an inter-frame difference encoding method encodes error information between frames. Was invented.

[0008]

In the above conventional technique, for example, for encoded data of a movie with subtitles,
Consider a case in which encoded data is thinned out only in a video portion while subtitles are left as they are. In the conventional method, since the entire frame is thinned, subtitles are thinned at the same time as the video is thinned, and essential information is lost. Conversely, in order to prevent subtitles from being decimated, frames without subtitles are not decimated, and only frames without subtitles are decimated, so that only image portions that cannot be watched have been decimated.

[0009] In addition, for a video such as a news program in which an announcer and a site video are in the same frame, for example, an announcer part is decimated, and a thinning process is performed for the purpose of not decimating the site video. There was a disadvantage that it could not be done.

[0010] As described above, the conventional thinning-out processing method limits the applicable video, and generally cannot be put to practical use when the video content differs for each region in the frame.

The present invention has been made in view of the above-described drawbacks and problems of the prior art, and provides an apparatus for generating encoded image data that enables frame decimation for each area depending on video content. The task is to

[0012]

In order to solve the above-mentioned problems, a first aspect of the present invention is to perform a frame thinning process on encoded image data which has been encoded in advance, and to newly add encoded image data. In the image encoding apparatus for creating the image encoding device, an area separating unit that classifies the image encoded data that has been encoded in advance into a first region and a second region, and a frame to be thinned out for the classified first encoded data. Frame decimating switch unit for classifying coded data of a frame that is not decimated from the coded data of frame No. 1, and a frame decimating calculation unit that performs a culling process of coded data of the first area classified as coded data of the frame to be culled And information indicating that the encoded data of the first area has been thinned and the encoded data of the second area, or the first
And a coded data merging unit for generating coded image data of a frame from the coded data of the region (b) and the coded data of the second region.

[0013] The encoded data merging section may be configured to execute the first data when the encoded data in the first area is thinned out.
A non-coding information generating unit that generates non-coding information indicating that the image coded data of the area is the same as the previous image, and from the non-coding information and the coded data of the second area, or A data merge unit that generates image encoded data of the frame from the encoded data of the first area and the encoded data of the second area of the frame that is not thinned out.

[0014] To solve the above-mentioned problem, a second invention of the present invention is an image coding apparatus for performing frame thinning processing on pre-encoded image encoded data to newly generate encoded image data. In the encoding device, an area separating unit that classifies image encoded data that has been encoded in advance into a plurality of areas, and does not thin out the encoded data of a frame to be thinned out for each of the encoded data of the plurality of areas that have been classified. A plurality of frame thinning-out switches for separating the encoded data of the frame, a plurality of frame thinning-out calculating units for thinning out the encoded data of each of the areas separated as the encoded data of the thinned-out frame, Encoded data for generating image encoded data of a frame from the thinned-out encoded data or the encoded data that is classified as not thinned out And over di unit, characterized in that it comprises.

Further, the area separating section separates the encoded image data which has been encoded in advance into a plurality of areas, and stores specific encoded data in an area other than the plurality of areas for each of the plurality of areas. A plurality of image data complementing units for merging, wherein the encoded data merging unit deletes specific encoded data of an area other than the area merged for each of the plurality of areas for each of the plurality of areas. Complementary data deletion part of,
A data merging unit that generates encoded image data by arranging the encoded data of each area in which the specific encoded data is deleted for each of the plurality of areas.

According to the first aspect of the present invention, it is possible to apply a conventional area separation technique using an image recognition technique or the like, and to perform frame thinning only in a specific area in accordance with the image content. For example, when transmitting a movie with subtitles, it is possible to realize frame skipping of only the video portion as follows.

First, the coded data in the frame is separated into a video area and a subtitle area by an area separating unit, and only the coded data of the video area is a frame to be decimated or not to be decimated by a frame decimation switch unit. If the frame is not thinned, it is transmitted to the coded data merge unit. If the frame is thinned, the frame is thinned out by the frame thinning calculation unit to create coded data, and the coded data merge unit is created. Send to The frame decimation unit performs frame decimation of the encoded data in the separated area using the conventional frame decimation process. The coded data merge unit uses the coded data of the caption separated by the region separation unit and the coded data of the video region obtained from the frame culling switch unit or the frame culling calculation unit to divide the coded data of one frame. create. In the case of a frame in which the video area is not thinned, the reference image is changed by the encoded data of the caption and the encoded data of the video area obtained from the frame thinning switch unit or the frame thinning obtained from the frame thinning calculation unit. One frame of encoded data is created using the encoded data of the image. On the other hand, in the case of a frame in which a video area is thinned out, one frame is obtained by using encoded data for a subtitle and using encoded data indicating that the image encoded data is the same as the previous image in the video area. Create encoded data of As described above, encoded data in which only the video portion of a movie with a balloon is thinned out can be created.

Further, according to the second aspect of the present invention, it is possible to perform frame thinning independently for each area according to the image content. For example, in an image of a news program in which an announcer and a site video exist, frame thinning of the announcer and the site video can be independently executed.

It is assumed that a frame thinning switch unit (1) and a frame thinning operation unit (1) for an announcer part, and a frame thinning switch unit (2) and a frame thinning operation unit (2) for an on-site video are provided. The region separating unit separates the encoded data in the frame into an announcer part and a site video. The coded data of the announcer part is discriminated in the frame decimating switch unit (1) depending on whether the frame is a decimated frame or a non-decimated frame. If the frame is not decimated, the coded data is transmitted to the coded data merging unit and decimated. In the case of a frame, the frame decimation unit (1) performs frame decimation to generate encoded data and transmits it to the encoded data merge unit. On the other hand, the coded data of the on-site video is discriminated in the frame decimating switch section (2) depending on whether the frame is a decimated frame or a non-decimated frame. If the frame is not decimated, the coded data is transmitted to the coded data merge section. In the case of a frame, the frame thinning operation unit (2) performs frame thinning to generate encoded data and transmits the encoded data to the encoded data merging unit. The coded data merging unit obtains the coded data of the announcer part obtained from the frame thinning switch unit (1) or the frame thinning operation unit (1), and obtains the coded data from the frame thinning switch unit (2) or the frame thinning operation unit (2). One frame of coded data is created using the coded data of the site video to be obtained. The frame thinning of the announcer part is performed by the frame thinning switch unit (1) and the frame thinning operation unit (1) independently of the encoded data of the local video, and the frame thinning of the local video is performed by the encoded data of the announcer part. Independently generate the encoded data of one frame by performing the operations in the frame thinning switch unit (2) and the frame thinning operation unit (2). As described above, it is possible to create encoded data in which the announcer portion and the on-site video are thinned out independently.

[0020]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, a first embodiment of the present invention will be described with reference to the drawings.

The video is a movie with a balloon, and in the present embodiment, H.264 is used as the encoding method. It is assumed that encoding is performed as follows using H.263. The size of the image is Q
CIF. A GOB having the first macroblock at the top of the frame and a GOB having the first macroblock at the bottom of the lowest macroblock line at the top is a GOB.
1. The GOB starting from the head macroblock of the lowest macroblock line is referred to as GOB2. AnnexR
Is applied to eliminate the need to refer to an image of a different GOB, so that encoding can be performed independently of other GOBs. The video part is in GOB1, the subtitle is in GOB2, and the video part is GOB2 depending on the time.
Or the subtitle part does not exist in GOB1. The video is 30 frames per second. It is assumed that only the first frame is an I-frame that has been intra-coded, and that all subsequent frames are P-frames using inter-frame differential coding using the previous frame as a reference image.

It is assumed that the area to be thinned out, the number of frames, and the number of frames are arbitrarily set in advance. In the present embodiment, the subtitles are set to 30 frames per second and the video portion is set to 15 frames per second for such encoded data. Do what you want. In addition, since it is assumed that the video part exists in GOB1 and the subtitle exists in GOB2, when separating the encoded data of the video part and the encoded data of the subtitle, it is determined by the GOB number. And executed.

FIG. 1 is a diagram showing the configuration of an embodiment of the first invention of the present invention.

In this embodiment, the area selection unit 101
Separates the input coded data into that of GOB1 and that of GOB2.

In the present embodiment, the frame thinning switch unit determines whether to output coded data as input data without performing frame thinning or to output as frame thinned encoded data. A determining unit 107, a frame thinning-out changing unit 102 for changing whether to output coded data as input data without performing frame thinning, or to output as frame-thinned encoded data, And a frame thinning count unit 103 for counting.

Further, in the present embodiment, the encoded data merge section creates encoded data absence information for creating encoded data indicating that the video information when the GOB1 portion is decoded is the same as the previous frame. And a GOB merging unit 106 that creates one frame of encoded data from the encoded data of GOB1 and the encoded data of GOB2.

Also, in the present embodiment, the frame thinning count unit 103 counts the number of frames of the encoded data, and uses a signal for causing the frame thinning calculation unit 104 to perform frame thinning when the target thinning number is reached. A certain frame thinning signal is transmitted.

The frame decimation operation unit 104 outputs the GOB1
It is assumed that the frame thinning of only the encoded data is performed. In the frame thinning calculation unit 104, frame thinning is performed by the same processing as that in the related art. In the present embodiment, the input coded data of GOB1 is stored, and the frame thinning signal is received from the frame thinning count unit 103. Then, using the accumulated encoded data, the frames are thinned by the number counted by the frame thinning count unit 103, and the encoded data of the last input frame is calculated.

In the frame thinning change means 102,
The change direction when the encoded data is not thinned out is point 1 and the change direction when the thinned frame is thinned out is point 2.

With the above arrangement, the first frame to the fifth frame
Before processing up to the frame, each unit operates as follows.

For the first frame, the input coded data is input to the area selector 101 by the data of GOB1 and the data of GOB2.
Separated into things. The frame thinning determination unit 107
In order to process the first frame as encoded data that does not perform frame thinning, the frame thinning changing unit 102
To point 1. The coded data of GOB1 is accumulated in the frame thinning-out operation unit 104. The encoded data of GOB1 and the encoded data of GOB2 are input to the GOB merge unit 106. Therefore, the output encoded data of the first frame is created and output.

For the second frame, the input coded data is separated by the area selection unit 101 into that of GOB1 and that of GOB2. The frame thinning determination unit 107 changes the frame thinning change unit 102 to point 2 in order to output the second frame as encoded data obtained by performing frame thinning. The coded data of GOB1 is accumulated in the frame thinning-out operation unit 104. The frame thinning count unit 103 counts 1 as the number of frames to be thinned. The coded data of GOB2 is input to the GOB merge unit 106. The coded data absence information creation unit 105
Encoded data indicating that the video information when the OB1 portion is decoded is the same as the previous frame is created and output to the GOB merging unit 106. The GOB merge unit 106 creates and outputs one frame of encoded data from the input encoded data of GOB1 and the encoded data of GOB2.

For the third frame, the input coded data is separated by the area selection unit 101 into that of GOB1 and that of GOB2. The frame thinning determination unit 107 changes the frame thinning change unit 102 to point 2 in order to output the third frame as encoded data obtained by performing frame thinning. The coded data of GOB1 is accumulated in the frame thinning-out operation unit 104. Since the target number of frames has been decimated in the previous frame, the frame culling count unit 103 transmits a frame culling signal to the frame culling calculation unit 104, and clears the counted number of culled frames to zero. Using the accumulated encoded data from the first frame to the third frame, the frame decimation unit 104 creates encoded data of a third frame using the first frame as a reference image. The created encoded data of GOB1 and the encoded data of GOB2 separated by the area selection unit 101 are input to the GOB merge unit 106. Therefore, output encoded data of the third frame is created and output.

For the fourth frame, the input coded data is separated by the area selection unit 101 into that of GOB1 and that of GOB2. The frame thinning determination unit 107 changes the frame thinning change unit 102 to point 2 in order to output the fourth frame as encoded data obtained by performing frame thinning. The coded data of GOB1 is accumulated in the frame thinning-out operation unit 104. The frame thinning count unit 103 counts 1 as the number of frames to be thinned. The coded data of GOB2 is input to the GOB merging unit. The coded data absence information creating unit 105 outputs
Coded data indicating that the video information obtained by decoding the portion is the same as the previous frame, and the GOB merging section 1
06 is output. The GOB merge unit 106 creates and outputs one frame of encoded data from the input encoded data of GOB1 and the encoded data of GOB2.

For the fifth frame, the input coded data is separated by the area selection unit 101 into that of GOB1 and that of GOB2. The frame thinning determination unit 107 changes the frame thinning change unit 102 to point 2 in order to output the fifth frame as encoded data obtained by performing frame thinning. The coded data of GOB1 is accumulated in the frame thinning-out operation unit 104. Since the target number of frames has been decimated in the previous frame, the frame culling count unit 103 transmits a frame culling signal to the frame culling calculation unit 104, and clears the counted number of culled frames to zero. Using the accumulated encoded data from the first frame to the fifth frame, the frame thinning calculation unit 104 creates encoded data of the fifth frame using the third frame as a reference image. The created encoded data of GOB1 and the encoded data of GOB2 separated by the area selection unit 101 are input to the GOB merge unit 106. Therefore, the output encoded data of the fifth frame is created and output.

In the present embodiment, when there are an area in which frame thinning is performed and a storage area in which no frame thinning is performed,
Frame thinning can be performed independently for only the region where the frame is to be performed.

For the sake of simplicity, the description has been given of the case where the size of the image is QCIF and the image of 30 frames per second is thinned out. However, the embodiment of FIG. 1 is not limited to this. Needless to say, the present invention can be applied to the number of frames and the thinning interval, and an effective thinning process can be performed. Further, in the example of the embodiment, H.264 is used. Although the H.263 coding method was used, the coding method is not limited to this, and it goes without saying that any coding method that generally performs coding in frame units is similarly applicable.

Next, a second embodiment of the present invention will be described with reference to the drawings.

The video is a news Banso video in which an announcer and a site video simultaneously exist. The encoding method is MPE
It is assumed that the encoding is performed as follows using G-2. The size of the video is QCIF. A portion of the site video is defined as a region 1 and a portion including an announcer other than the site video is defined as a region 2. The region 1 and the region 2 are distinguished on a macroblock basis, and the positions and shapes in the drawing do not change with time without referring to each other in the inter-frame differential encoding method. It is assumed that the area separation uses a general image recognition technique. Corresponding to the two regions, it is assumed that two frame thinning switch units and two frame thinning calculation units are provided. The video is 30 frames per second. Only the first frame is an intra-coded I frame, the odd-numbered frames thereafter are P frames using inter-frame differential coding with the previous frame as a reference image, and the even-numbered frames are It is assumed that the subsequent frame is a B-frame using the inter-frame difference encoding with the reference image.

In such coded data, in area 1, all B frames are thinned out to 15 frames per second, and in area 2, all B frames and some P frames are thinned out. To make 7.5 frames per second.

FIG. 2 is a diagram showing a configuration of an embodiment of the second invention of the present invention.

In the present embodiment, the area selecting section includes a data separating section 201 for separating the encoded data into the encoded data of the area 1 and the encoded data of the area 2, and the RGB of the image of the area 2. Image data interpolator (1) 202 for merging data encoded as 0 with encoded data in area 1
And an image data interpolator (2) 208 for merging data obtained by setting the RGB of the image of the area 1 to 0 with the encoded data of the area 2.

The frame thinning switch unit for the area 1
A frame thinning-out unit (1) 203 for determining whether to output coded data as input data without performing frame thinning or as coded data obtained by performing frame thinning, and to perform frame thinning on the coded data; Frame thinning-out changing means (1) 204 for changing whether to output the input data without performing it or to output it as encoded data obtained by performing frame thinning.

The frame thinning-out switch unit for the area 2
A frame thinning determination unit (2) 209 for determining whether to output the encoded data as input data without performing frame thinning or to output the data as frame-thinned encoded data; A frame thinning change unit (2) 210 for changing whether to output the input data without performing it or to output it as encoded data obtained by performing frame thinning, and a frame thinning count unit 211 for counting the number of thinned frames. Be composed.

The encoded data merging section includes an interpolation data deleting section (1) 207 for deleting encoded data corresponding to the area 2 from the encoded data obtained from the frame thinning switch section for the area 1; Area 1 of the encoded data obtained from the frame thinning switch unit for
And a data merging unit 205 for creating one frame of encoded data from the encoded data of the area 1 and the encoded data of the area 2. Is done.

In the present embodiment, the frame thinning counting section 211 counts the number of frames of the encoded data, and when the number of frames reaches the target thinning number, causes the frame thinning calculating section (2) 212 to perform frame thinning. A frame thinning signal, which is a signal, is transmitted.

The frame thinning-out unit (1) 206 deletes the input coded data and creates coded data indicating the same image information as the previous frame. On the other hand,
The frame thinning calculation unit (2) 212 accumulates the input encoded data and stores the encoded data.
When a frame thinning signal is received from 11, the encoded data of the last input frame is calculated using the accumulated input data as a reference image with the first input frame as a reference image.

In the frame thinning-out changing means (1) 203, the change direction in which encoded data is not thinned out is point 1, and the change direction in case of frame thinning out is point 2. In the frame thinning change unit (2) 210, a change direction when encoding data is not thinned is set to point 3, and a change direction when frame thinning is set to point 4.

Also, since the target of frame thinning for the area 1 is only the B frame, the frame thinning can be realized only by deleting the encoded data of the corresponding B frame. On the other hand, since the target of the frame thinning for the area 2 is the B frame and the P frame, in order to perform the frame thinning, the encoded data of all the frames from the encoded data of the first frame is accumulated. For this reason, the area 2 is provided with a frame thinning-out count unit 211 for determining a frame to be thinned out.

The interpolated data deleting unit (1) 207 and the interpolated data deleting unit (2) 213 delete the encoded data of the macroblock whose RGB is set to 0, so that the encoded data of the area 1 and the encoded data of the area 2 are deleted. The encoded data is determined, and the encoded data is deleted. The data merge unit 205 includes an interpolation data deletion unit (1) 207 and an interpolation data deletion unit (2) 21.
In step 3, the coded data of the macro blocks that are not deleted are sequentially arranged to create coded data of one frame.

Based on the above configuration, the first frame through the fifth frame
Before processing up to the frame, each unit operates as follows.

For the first frame, the input coded data is separated by the data separation section 201 into coded data into a region 1 portion and a region 2 portion. The area 1 is transmitted to the image data interpolator (1) 202, and the area 2 is transmitted to the image data interpolator (2) 208. The image data interpolation unit (1) 202 creates encoded data for one frame by interpolating the encoded data of the area 2 portion with data obtained by encoding image information in which all RGB are 0.
The frame thinning determination unit (1) 203 changes the frame thinning change unit (1) 204 to point 1 in order to process the first frame as encoded data without performing frame thinning. Next, the interpolation data deletion unit (1) 2
At 07, the encoded data in the area 2 is deleted. On the other hand, the image data interpolating unit (2) 208 interpolates the encoded data of the area 1 with the data obtained by encoding the image information of which RGB is all 0 to generate encoded data for one frame. The frame thinning determination unit (2) 209 converts the first frame as coded data that does not perform frame thinning.
To point 3. Image data interpolation unit (2) 20
The encoded data created in step 8 is stored in the frame thinning out section (2) 212. Then, the coded data of the area 1 is deleted by the interpolation data deletion unit (2) 213. The encoded data of the area 1 is input to the data merging section 205 from the interpolation data deleting section (1) 207 and the encoded data of the area 2 from the interpolation data deleting section (2) 213. The data merging unit 205 creates and outputs one frame of encoded data from the input encoded data.

For the second frame, the input coded data is separated by the data separation section 201 into coded data into a region 1 portion and a region 2 portion. The area 1 is transmitted to the image data interpolator (1) 202, and the area 2 is transmitted to the image data interpolator (2) 208. The image data interpolation unit (1) 202 creates encoded data for one frame by interpolating the encoded data of the area 2 portion with data obtained by encoding image information in which all RGB are 0. The frame thinning determination unit (1) 203 converts the frame thinning change unit (1) 20 to thin the second frame.
Change 4 to point 2. The coded data is input to the frame thinning operation unit (1) 206. The frame decimation unit (1) 206 deletes the input B frame,
The coded data indicating that the image information is the same as the previous frame is transmitted to the interpolation data deletion unit (1) 207. Then, the coded data of the area 2 is deleted by the interpolation data deletion unit (1) 207. On the other hand, the image data interpolation unit (2) 208 converts the encoded data of the area 1 into RGB data.
Are interpolated by encoded data of image information in which all are 0, thereby generating encoded data for one frame. The frame thinning determination unit (2) 209 changes the frame thinning change unit (2) 210 to point 4 in order to output the second frame as the frame thinned encoded data. The encoded data created by the image data interpolation unit (2) 208 is accumulated in the frame thinning out calculation unit (2) 212. The frame thinning count unit 211 counts 1 as the number of frames to be thinned. The frame thinning-out unit (2) 212 transmits encoded data indicating that the image information is the same as the previous frame to the interpolation data deleting unit (2) 213. Then, the coded data of the area 1 is deleted by the interpolation data deletion unit (2) 213. The coded data of the area 1 is input from the interpolation data deletion unit (1) 207 to the data merge unit 205 from the interpolation data deletion unit (2) 213. Data merge unit 2
Reference numeral 05 creates and outputs one frame of encoded data from the input encoded data.

With respect to the third frame, the input coded data is separated from the coded data into a region 1 and a region 2 by a data separation unit. The area 1 is transmitted to the image data interpolator (1) 202, and the area 2 is transmitted to the image data interpolator (2) 208. The image data interpolation unit (1) 202 creates encoded data for one frame by interpolating the encoded data of the area 2 portion with data obtained by encoding image information in which all RGB are 0. The frame thinning determination unit (1) 203 sets the frame thinning change unit (1) 204 to point 1 in order to process the third frame as encoded data that does not perform frame thinning.
Change to Next, the interpolation data deletion unit (1) 207
The coded data in the area 2 is deleted. On the other hand,
The image data interpolation unit (2) 208 interpolates the encoded data of the area 1 with data obtained by encoding image information in which RGB is all 0 to create encoded data for one frame. The frame thinning determination unit (2) 209 changes the frame thinning change unit (2) 210 to point 4 in order to output the eighth frame as encoded data obtained by performing frame thinning. The encoded data created by the plane image data interpolation unit (2) 208 is stored in the frame thinning calculation unit (2) 212. The frame thinning count unit 211 counts 1 as the number of frames to be thinned. The frame thinning-out unit (2) 212 transmits encoded data indicating that the image information is the same as the previous frame to the interpolation data deleting unit (2) 213. Then, the coded data of the area 1 is deleted by the interpolation data deletion unit (2) 213. The coded data of the area 1 from the interpolation data deletion unit (1) 207 is transmitted from the interpolation data deletion unit (2) 213 to the area 2
Is input to the data merging unit 205.
The data merging unit 205 determines 1 from the input encoded data.
Create and output encoded data for the frame.

With respect to the fourth frame, the input coded data is separated from the coded data into a region 1 and a region 2 by a data separation unit. The area 1 is transmitted to the image data interpolator (1) 202, and the area 2 is transmitted to the image data interpolator (2) 208. The image data interpolation unit (1) 202 creates encoded data for one frame by interpolating the encoded data of the area 2 portion with data obtained by encoding image information in which all RGB are 0. The frame thinning determination unit (1) 203 changes the frame thinning change unit (1) 204 to point 2 in order to thin the fourth frame. The coded data is input to the frame thinning operation unit (1) 206. The frame thinning-out unit (1) 206 deletes the input B frame, and transmits coded data indicating the same image information as the previous frame to the interpolation data deleting unit (1) 207. Then, the coded data in the area 2 is deleted by the interpolation data deletion unit (1) 207. On the other hand, the image data interpolation unit (2) 2
A step 08 interpolates the coded data in the area 1 with RGB or data obtained by coding image information that is all 0 to create coded data for one frame. The frame thinning determination unit (2) 209 changes the frame thinning change unit (2) 210 to point 4 in order to output the fourth frame as encoded data subjected to frame thinning. The encoded data created by the image data interpolation unit (2) 208 is accumulated in the frame thinning out calculation unit (2) 212.
The frame thinning count unit 211 counts 1 as the number of frames to be thinned. Frame thinning calculation unit (2) 21
2 transmits to the interpolation data deletion unit (2) 213 encoded data indicating that the image information is the same as the previous frame. Then, the coded data of the area 1 is deleted by the interpolation data deletion unit (2) 213. The encoded data of the area 1 is input to the data merging section 205 from the interpolation data deleting section (1) 207 and the encoded data of the area 2 from the interpolation data deleting section (2) 213. The data merging unit 205 creates and outputs one frame of encoded data from the input encoded data.

For the fifth frame, the input coded data is separated by a data separation unit into coded data into a region 1 portion and a region 2 portion. The area 1 is transmitted to the image data interpolator (1) 202, and the area 2 is transmitted to the image data interpolator (2) 208. The image data interpolation unit (1) 202 creates encoded data for one frame by interpolating the encoded data of the area 2 portion with data obtained by encoding image information in which all RGB are 0. The frame thinning determination unit (1) 203 sets the frame thinning change unit (1) 204 to point 1 in order to process the fifth frame as encoded data that does not perform frame thinning.
Change to Next, the interpolation data deletion unit (1) 207
The coded data in the area 2 is deleted. Other power,
The image data interpolation unit (2) 208 interpolates the encoded data of the area 1 with data obtained by encoding image information in which RGB is all 0 to create encoded data for one frame. The frame thinning determination unit (2) 209 changes the frame thinning change unit (2) 210 to point 4 in order to output the fifth frame as frame thinned encoded data. The encoded data created by the image data interpolation unit (2) 208 is accumulated in the frame thinning out calculation unit (2) 212. Since the target number of frames has been decimated in the previous frame, the frame culling count unit 211 transmits a frame culling signal to the frame culling calculation unit (2) 212, and clears the count of culled frames to zero. The frame thinning-out unit (2) 212 uses the accumulated encoded data from the first frame to the fifth frame to create encoded data of the fifth frame using the first frame as a reference image. Then, the coded data of the area 1 is deleted by the interpolation data deletion unit (2) 213. Area 1 from interpolation data deletion unit (1) 207
Is input to the data merging unit 202 from the interpolation data deleting unit (2) 213. The data merging unit 205 creates and outputs one frame of encoded data from the input encoded data. In the present embodiment, when there are a plurality of regions where frame thinning is performed at different intervals in a frame, frame thinning can be performed for each region.

In the above, for the sake of simplicity, the description has been given of the case where the image size is QCIF and the image of 30 frames per second is thinned out. However, the embodiment of FIG. 2 is not limited to this. Needless to say, the present invention can be applied to the number of frames and the thinning interval, and an effective thinning process can be performed. Furthermore, in the embodiment, MPEG-2
Although the encoding system is used, the encoding system is not limited to this, and it goes without saying that any encoding system that generally performs encoding in frame units can be similarly applied.

[0058]

According to the conventional method, frame thinning is performed for the entire frame. Therefore, it is impossible to perform frame thinning at different intervals for each area in the screen. However, according to the present invention, the coded data is once separated for each image region and the frame thinning process is performed independently according to the image content. As a result, it is possible to create encoded data in which frame thinning is performed independently for each region.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of an image coded data creation device according to an embodiment of the first invention of the present invention.

FIG. 2 is a diagram showing a configuration of an image coded data creation device according to an embodiment of the second invention of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 101 ... Area selection part 102 ... Frame thinning change means 103 ... Frame thinning count part 104 ... Frame thinning calculation part 105 ... Encoded data absence information creation part 106 ... GOB merge part 107 ... Frame thinning determination part 201 ... Data separation part 202 ... Image data interpolation unit (1) 203 ... Frame thinning determination unit (1) 204 ... Frame thinning change unit (1) 205 ... Data merge unit 206 ... Frame thinning calculation unit (1) 207 ... Interpolation data deletion unit (1) 208 ... Image data interpolation unit (2) 209: frame thinning determination unit (2) 210: frame thinning change unit (2) 211: frame thinning count unit 212: frame thinning calculation unit (2) 213: interpolation data deletion unit (2)

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Susumu Ichinose 3-19-2 Nishi-Shinjuku, Shinjuku-ku, Tokyo F-term in Nippon Telegraph and Telephone Corporation (reference) 5C023 AA27 BA15 CA01 CA08 5C059 KK39 LB07 LB13 MA00 PP04 RC12 SS06 UA02

Claims (4)

[Claims] 1. An image coding apparatus for performing frame thinning processing on previously encoded image encoded data to newly generate image encoded data, comprising the steps of: Area and the second
An area separating unit that separates the first encoded data into a coded data of a frame to be decimated and a coded data of a frame that is not decimated; The first that has been classified as encoded data
A frame thinning-out operation unit for thinning out the encoded data of the area, and information indicating that the encoded data of the first area has been thinned and the encoded data of the second area, or Coded image data comprising: a coded data merging unit for generating coded image data of a frame from coded data of a first area of the existing frame and coded data of the second area. Creating device. 2. The encoding data merging unit according to claim 1, wherein when the encoded data in the first area is thinned out, the encoded data merger indicates that the encoded image data in the first area is the same as the previous image. A non-coding information generating unit for generating non-coding information; and the non-coding information and the coding data of the second area, or the coding data of the first area of the undecimated frame and the second data. The apparatus according to claim 1, further comprising: a data merging unit configured to generate image encoded data of a frame from the encoded data of the area. 3. An image coding apparatus for performing frame decimation processing on pre-encoded image encoded data and newly creating image encoded data, comprising the steps of: An area separating unit that separates the data into regions; a plurality of frame thinning switch units that separate the encoded data of the plurality of separated regions into coded data of a thinned frame and coded data of a non-thinned frame; A plurality of frame thinning-out operation units for thinning out the coded data of each of the areas separated as the coded data of the thinned-out frame; coded data thinned out for each of the areas or a code separated without the thinning-out And a coded data merging unit for generating coded image data of the frame from the coded data. Creation device. 4. A data separation unit for separating image-encoded data encoded in advance into a plurality of regions, wherein the region separation unit separates specific encoded data into regions other than the plurality of regions for each of the plurality of regions. A plurality of image data complementing sections to be merged, wherein the encoded data merging section deletes specific encoded data of an area other than the area merged for each of the plurality of areas for each of the plurality of areas. And a data merging unit that generates encoded image data by arranging the encoded data of each area in which the specific encoded data has been deleted for each of the plurality of areas. 4. The apparatus according to claim 3, wherein: