JP3296758B2 - Moving picture coding apparatus and moving picture coding / reproducing apparatus - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coding apparatus for coding a moving picture, and more particularly to a moving picture suitable for application to a video rate compressing and reproducing apparatus such as a video recorder or an image transmitting and receiving apparatus. code KaSo 置及 beauty video encoding / playback apparatus.

[0002]

2. Description of the Related Art It is an object of the present invention to store image information on a digital video disk (DVD) or the like and to use it for a video coding method of digital broadcasting, etc. by a high efficiency coding process applying a digital image compression technique. ISO / IE
MP @ ML (Mai) in MPEG-2 standardized by C
n Profile, Main Level), the PAL method uses a spatial resolution
720 x 576 pixels in the horizontal and vertical directions (NT
In the case of the SC system, 720 × 480 pixels), the bit rate is a maximum of 15 Mbit / sec, and the frame rate is a maximum of 25 frames / sec (in the case of the NTSC system, 29.97.
(Frames / second).

An encoding device for encoding a moving image based on this standard has been proposed. For example, in the 1997 IEICE General Conference C-12-33-37, real-time encoding / coding was performed using a one-chip LSI and an external memory.
The case where decoding is realized is shown.

[0004] Although the moving picture coding apparatus is disclosed in Japanese Patent Laid-Open No. 5-236466, which will be described with reference to FIG. Here, 1 is an image information input terminal into which moving image information is input in the display order of images, 202 is a frame order rearrangement control circuit for rearranging the frame order of the input image in accordance with the image sequence and outputting the image in the encoding order of the images, 2
03 is a storage medium for holding and outputting an input image in accordance with the frame order rearrangement control circuit 202 , 204 is a motion detection circuit for a P image (forward prediction coded image) or B image (bidirectional prediction coded image), and 205 is a storage. Medium control circuit, 206 is a storage medium, 207 is a predicted image generation circuit, 2
08 is an encoded image determination circuit, 209 is an encoded image selection circuit, 210 is a DCT (Discrete Cosine Transform) circuit, 211 is a quantization circuit, 212 is a variable length encoding circuit, and 214 is an inverse quantization circuit 215 is I
DCT (Inverse Discrete Cosine Transform) circuit, 216 is a decoded image generation circuit,
Reference numeral 101 denotes a storage medium, and reference numeral 17 denotes an encoded information output terminal that outputs encoded information for an encoded image.

[0005] Here, a GOP (Group of Pictures) is constituted by a plurality of continuous frames, and the types of images constituting the GOP are usually as follows.
A forward coded image using forward motion prediction on the time axis, using an intra-coded image (I image) coded only with its own image information, a past I image or P image as a reference image. Predictively coded image (P image), bidirectional predictive coded image (B image) coded using forward and backward motion prediction on the time axis using past and future I images or P images as reference images Are classified into three types.

[0006] The structure of the GOP is such that classification is performed according to the M value determined by the number of frames of continuous B images. For example, when M = 3, as shown in FIG. 2 and FIG. B image, B image,
The I image, the B image, the B image, the P image,..., And the B image are continuously generated in two frames. The input moving image information is obtained by determining a block composed of 8 × 8 pixels and a macro block composed of 16 × 16 pixels to be subjected to motion vector detection as an encoding target and performing predictive encoding. It is.

In FIG. 13, a frame order rearranging control circuit 202 stores input image information 1a to be input in a display order according to an M value in a storage medium 203, and encodes the input image information 1a in a coding frame order and a macroblock unit by a block scan. And reads it out and outputs it as input image information 203a.

[0008] In the storage medium 206, the local decoded image information 216a for the I image or the P image encoded in the most recent past and reproduced by the local decoding processing is
In accordance with the storage medium control circuit 205, the data is stored and stored, and when a P image or a B image is encoded, it is reproduced and output as a reference image for a motion search at a subsequent stage of the motion detection circuit 204.

The P image and the B image are
At the time of predictive encoding of an image, the decoded image information of the past I image or P image is read out from the storage medium 206 as a reference image for each macroblock, and a motion vector is detected in multiple stages. At the same time, the input image information is output 204a. In this case, the image information for the search area is accessed from the storage medium 206 for each motion search in each stage. By the storage medium control circuit 5,
The storage medium 206 is accessed according to the motion vector information, and the predicted image information 206a is read.

The input image information 204a and the input image information 2
06a to obtain a difference image and predictive difference image information 207a
Is generated, and the input image information 207b and the predicted image information 207 are generated.
c is output.

The coded image determination circuit 208 and the coded image selection circuit 209 compare the prediction difference image 207a (non-intra macro block) and the input image 207b (intra macro block) for each macro block.
An image having a smaller code amount when each is encoded is determined as an encoded image, and is output as encoded image information 209a.

The encoded image information 209a is a DCT
The circuit 210 performs discrete cosine transform on a block basis, and the quantized circuit 211 quantizes the transform coefficient to obtain quantized information 211a. This quantization information 211a is
The variable length coding circuit 212 codes the motion vector information and the prediction mode together with a bit sequence, and outputs the bit sequence as coded information 212a.

On the other hand, the quantized information 211a is reproduced as predictive difference image information 215a by an inverse quantization circuit 214 and an IDCT circuit 215, and the decoded image generation circuit 216 outputs an I image or P image in macroblock units. At the time of local decoding processing, the prediction difference image information 215a is used for an intra macroblock, and the prediction image generation circuit 207 is used for a non-intra macroblock.
Are added to the prediction difference image information 215a to the prediction image information 207c generated by the local decoding image information 2
16a, and is stored in the storage medium 206. Further, at the time of the local decoding processing of the B image, the local decoded image information 216a is stored and held in the storage medium 101.

The decoded image information 206b of the I image and the P image stored and held in the storage medium 206 by the storage medium control circuit 205 in the order of displaying the frames of the images and in the order of scanning.
The decoded image information 101a of the B image stored and held in the storage medium 101 is read out and output to the monitor display circuit 220a via the monitor output circuit 220.

Here, the storage medium 20 is used for the multi-step search.
6 needs to be accessed at each stage,
In the search at each stage, access to the storage medium 206 is reduced in accordance with the subsample by using only the subsampled signal. Further, at least at the final stage, a motion search with half-pixel accuracy is performed. Since it is necessary to read out all the pixel information in the range and the generation of the predicted image requires the pixel information of all the pixels, the access for the motion search at the final stage and the access for creating the predicted image are performed. Are performed at the same time, thereby reducing the access frequency and improving the efficiency of memory access.
Also, when implementing this moving picture coding apparatus, instead of using memories for a plurality of chipsets and chips individually for each application, one LSI and a memory for each application are integrated into one (for example, a frame memory). By doing so, the power consumption of the moving picture coding apparatus is suppressed.

Furthermore, when coding a moving picture, the quantization information 211a for all frames, the station by the local decoding process
The partial decoded image information 216a is generated, and the decoded image information 206b or 101a is displayed on a monitor table in the order in which the images are displayed.
By outputting a monitor display as the indication information 220a ,
The encoding processing operation can be sequentially confirmed.

Although not shown, the moving picture coding apparatus having the above-described circuit configuration is provided with a variable length decoding circuit for performing processing in the reverse order of the variable length coding circuit 212, so that an externally reproduced coding can be performed. The information is input to a variable length decoding circuit (not shown), and the decoded image information is displayed on a monitor by a process similar to the local decoding process according to the quantization information obtained by the variable length decoding process and the coding conditions. Thus, a decoding device is realized by diverting the circuit blocks used for the above-described encoding processing.

[0018]

However, in a moving picture coding processing apparatus composed of the above-described one-chip LSI and an external memory (frame memory) integrated into one, the number of accesses is increased by increasing the efficiency of memory access. Even if the method of reducing power consumption is reduced, encoding of a moving image in real time, generation of locally decoded image information by local decoding, and monitor display output of the decoded image information Is performed by increasing the access bandwidth of the frame memory and increasing the operating frequency of the memory access, thus increasing the power consumption.

In addition, it is necessary to separately prepare a frame memory area for storing and holding decoded image information for display output even for a B image which is not used as a reference image at the time of encoding. Is inevitable.

An object of the present invention is to solve the above problem by performing encoding of a moving image in real time, instead of constantly operating all the circuit blocks used for the encoding. Depending on the type of image to be encoded, the operation of the circuit blocks that are not the object of the encoding process is stopped in a timely manner, and each circuit block is efficiently operated, so that unnecessary power consumption is suppressed and the necessary encoding Low power consumption and low cost video encoding apparatus and video encoding capable of sequentially monitoring the encoding process status even with a minimum memory capacity
/ Reproducing apparatus , especially suitable for video rate compression and reproduction of stationary or battery type portable video recorders, image transmitting and receiving apparatuses, etc., which can extend the usable time. It will be.

[0021]

In order to achieve the above-mentioned object, a first moving picture coding apparatus according to the present invention uses an input moving picture information as an I picture, a P picture and a B picture, and
In a moving picture coding apparatus that performs coding according to the G-2 method, a first storage medium that stores and holds input moving picture information and a control that controls the first storage medium and rearranges the input picture information in the coding order A frame order rearrangement control unit that outputs a local decoding processing permission signal, a second storage medium that stores and holds local decoded image information described later, and an I image and a P image that are used as reference images during encoding. A third storage medium that stores and holds input image information, a storage medium control unit that controls writing and reading of the second and third storage media, and P
At the time of encoding an image and a B image, motion detection is performed in multiple stages in each prediction direction to detect a motion vector with half-pixel accuracy and to specify a prediction mode. A predicted image generating unit that generates and outputs predicted image information indicated by a vector and a predicted difference image from the input image information; and codes one of the predicted difference image (non-intra macro block) and the input image (intra macro block). Image determining means and encoded image selecting means for determining encoded image information and outputting encoded image information, DCT means for performing discrete cosine transform of the encoded image information in block units, and quantizing transform coefficients. And a variable length encoding process of the quantization information obtained by the quantization unit together with the motion vector and the prediction mode, and Variable length encoding means for outputting encoded information, inverse quantization means for inversely quantizing the quantized information, IDCT means for decoding predicted difference image information by inverse discrete cosine transform, and local decoding of input image At the time of encoding, the decoded prediction difference image information is added to the decoded prediction difference image information in the case of an intra macroblock, or the decoded prediction difference image information is added to the decoded prediction difference image information in the case of a non-intra macroblock. Decoded image generating means for respectively obtaining local decoded image information by means of a local decoding processing permission signal output from the frame order rearranging control means. It controls the operation with the generation means, and when encoding the input image, in the case of a B image in which the decoded image information is not used as a reference image for predictive encoding of another image. Stops the operation of the decoded picture generation means and inverse quantization means and IDCT means.

According to this configuration, as a coding method for input moving image information, the type of the coded image is determined by the M value. For example, when M = 3, for example,
As shown in FIG. 4, the relationship between the input image frame and the type of the coded image is as follows: F0 (B image), F1 (B image), F2 (I image), F3 (B image), F4 (B ), F5 (P image),..., And the coding frame order is F2 (I image), F0 (B image), F1 (B image).
), F5 (P image), F3 (B image), F4 (B image),..., So that at least three frames are stored and held in the first storage medium, and at least three frames are stored in the second storage medium. The decoded image information for two frames is stored and held. In the third storage medium, the input image information corresponding to the I image or the P image is scan-converted by the storage medium control means and at least two frames are stored and held. At the time of predictive encoding of the P image and the B image, the input image information of the past I image or P image is read out from the third storage medium as a reference image for each macroblock by the motion detecting means as forward prediction. First stage motion vector is detected by search,
In accordance with the first-stage motion vector, decoded image information of a past I image or P image is read from the second storage medium, image information with half-pixel accuracy is generated, and the obtained image information is used as a reference image. The search detects a motion vector with half-pixel accuracy, which is the search accuracy of MPEG-2 , and specifies the prediction mode.

The prediction image generation means obtains a difference image between the input image information and the prediction image information in macroblock units, generates prediction difference image information, and outputs the input image information and the prediction image information, respectively. The coded image determination unit and the coded image selection unit compare the prediction difference image (non-intra macro block) with the input image (intra macro block), determine an image having a small code amount as the coded image, Output the encoded image information. The coded image information is obtained as quantization information by DCT means and quantization means, coded into a bit sequence together with motion vector information and a prediction mode by variable length coding means, and outputs coded information. On the other hand, the quantization information is obtained by the inverse quantization means and I
The prediction difference image is reproduced by the DCT means. The decoded image generating means outputs the locally decoded image information of the I image or the P image and stores the information in the second storage medium.

Since the locally decoded image information is not used as a reference image for predictive encoding of another image when encoding the B image, the local decoding processing permission of the frame order rearrangement control means is not used. Operation of the above-mentioned inverse quantization means and IDCT means for performing local decoding processing by a signal, addition processing of a predicted image and a predicted difference image in a decoded image generation means, output of locally decoded image information, storage medium control means and It is possible to stop the process of writing the locally decoded image information in the second storage medium. As a result, it is possible to stop the operation of the circuit blocks not included in the encoding process in a timely manner based on the characteristics of the encoded image instead of constantly operating all the circuits prepared for the encoding process of the moving image. And the encoding device is, for example, a one-chip LSI
In the case of moving picture encoding processing,
Each circuit block can be operated efficiently, and unnecessary power consumption can be suppressed.

Further, the second moving picture coding apparatus according to the present invention is different from the first moving picture coding apparatus in that the first storage medium for storing and holding input moving picture information, A monitor output control means for reading and controlling the second storage medium for storing and holding the locally decoded image information of the image, and outputting the input image information and the locally decoded image information in the display frame order and the scanning order, Monitor output selecting means for selecting one of the image information and the locally decoded image information and outputting the information on a monitor; and for the I image and the P image, the local decoding stored and held in the second storage medium. For the B image, the input image information stored and held in the first storage medium is output on the monitor in the display order.

According to this configuration, the input image information stored and held in the first storage medium prepared for rearranging the frame order of the input image in the order of the encoding frame, and the reference image for performing the predictive encoding The locally output image information of the I image and the P image stored and held in the second storage medium prepared for storing the information is converted by the monitor output control means according to the type of the coded image into the B image. the input moving image data, in the case of I picture or P picture, the local
The decoded image information is read out in the display frame order and in the scanning order, and the monitor output selection means determines whether the display image order is the locally decoded image information of the second storage medium or the input image information of the first storage medium. Select and output to monitor display.
Thus, a continuous and smooth monitor display output is obtained in real time.

For the B image, since the input image information is selectively output, for example, when M = 3, the decoded image information of the B image is stored in the second storage medium as compared with the case where the decoded image information of the B image is output on a monitor. as the capacity of a frame memory, it is possible to the memory capacity reduction of at least two frames, and the generation process of the local decoded image, since it is possible rather eliminates the process of writing to the frame memory, the access bandwidth of the frame memory Can be easily achieved without increasing the power consumption.

Further, in the moving picture coding / reproducing apparatus according to the present invention, in the first or second moving picture coding apparatus, a processing mode in which one of coding processing and decoding processing is instructed as an operation condition. , A coding mode input terminal for inputting playback coding information, a variable length decoding process for performing variable length decoding of the playback coding information, and playing back quantization information, a motion vector, and a prediction mode. The quantization means, as the quantization information to be decoded, either the output of the quantization means, or the output of the variable length decoding means,
Selecting means for selecting according to the processing mode;
A fourth storage medium in which the storage media of the first to third storage media are integrated into one and controlled by a common control signal; control signal selecting means for selecting a control signal of the fourth storage medium; and a fourth storage medium Data selection means for selecting write / read data of the frame order, wherein the frame order rearrangement control means outputs the local decoding processing permission signal and the encoding processing permission signal in accordance with the processing mode. The local decoding processing permission signal also controls the operation of the variable length decoding means, and the coding processing permission signal controls the motion detection means, the predicted image generation means, the coded image determination means, and the coded image selection. Means, DCT means, quantizing means, and variable-length coding means, and further, in the area of the fourth storage medium, the area corresponding to the first storage medium is not used at the time of decoding. Since assigns the decoded image information for B image in the area for storing and holding.

According to this configuration, the variable-length decoding means performs variable-length decoding on the externally encoded information, reproduces the quantization information, the motion vector information, and the prediction mode. The quantization information is decoded by the IDCT means to obtain prediction difference image information. Also, the reproduced motion vector information and the prediction mode are given to the motion detection means and the prediction image generation means, and the storage medium control means converts the prediction image information corresponding to the motion vector from the second storage medium according to the instruction of the motion detection means. The readout / predicted image generating means generates and outputs predicted image information obtained by interpolating past and future predicted image information when predicting image information passed from the motion detecting means or performing bidirectional prediction using a B image. The decoded image generation means adds the predicted image information and the predicted difference image information in accordance with the prediction mode, generates decoded image information, and stores the decoded image information in the fourth storage medium.

Here, as the decoded image information for the B image, an area corresponding to the first storage medium not used in the decoding process is allocated. Also, according to the local decoding processing permission signal and the coding processing permission signal of the frame order rearrangement control means, the variable length decoding means is encoded at the time of encoding, and the IDCT means and the inverse quantization means are decoded at the time of encoding the B image. A decoded image generating means, a motion search processing in the motion detecting means, a difference image generating processing in the predicted image generating means, a coded image determining means, a coded image selecting means,
When decoding the I image, the CT means, the quantization means, and the variable length coding means can further stop the operations of the motion detection means and the predicted image generation means.

As described above, it is possible to configure an encoding apparatus having a decoding process without using a circuit block prepared for encoding and adding a memory area by using a memory area.

Even in the case where the one-chip LSI and the external memory are separated from each other, useless external memory access is stopped in accordance with the type of the coded image, so that the data line and the control signal line are used. In addition to reducing power consumption, it is possible to easily reduce power supply noise and radiation noise.

Even when a coding device having a decoding process is constituted by a one-chip LSI, all the circuit blocks prepared for the coding process and the decoding process are always operated. Instead, according to the type of image to be coded or decoded, the operation of the circuit blocks that are not subject to the encoding or decoding processing is stopped in a timely manner so that each circuit block operates efficiently, thereby reducing wasteful power consumption. It is possible to achieve a low cost, low power consumption and low cost moving picture coding apparatus.

[0034]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a block diagram showing a first embodiment of a moving picture encoding apparatus according to the present invention, a decoded picture monitoring method and a control method using the same, and 1 is picture information inputted in display order. The input terminal 2 has a frame order rearrangement control circuit for rearranging the frame order of the input image according to the image sequence in the order of encoding of the image, and outputs the rearranged image. A first storage medium, 4 is a motion detection circuit for a P image or a B image, 5 is a storage medium control circuit, 6 is a second storage medium, 7 is a predicted image generation circuit, 8 is an encoded image determination circuit, 9 Is a coded image selection circuit, and 10 is a DCT (Discrete
te Cosine Transform) circuit, 11 is a quantization circuit, 1
2 is a variable length coding circuit, 13 is a third storage medium, 14 is an inverse quantization circuit, and 15 is an IDCT (Inverse Discrete Cosi
ne Transform) circuit, 16 is a decoded image generation circuit, 17
Is an encoded information output terminal for outputting encoded information for the encoded image, 18 is a monitor output circuit, and 19 is a monitor output terminal.

In this embodiment, in encoding and decoding of a moving image based on the MPEG-2 standard, MP @ ML (M
ain Profile, Main Level), and as described above, in the case of the PAL system, the spatial resolution is up to 7 in the horizontal and vertical directions as described above.
20 × 576 pixels (720 × in the case of NTSC system)
480 pixels), and handles an image (4: 2: 0 format) obtained by sub-sampling a color difference signal of the CCIR 601 standard (4: 2: 2 format) in a vertical direction of 2: 1 as an image format. Therefore, in the case of the PAL system, the luminance signal has a maximum of 720 × 576 pixels (in the case of the NTSC system, 720 × 480 pixels), and the color difference signals are PA signals, respectively.
In the case of the L system, encoding is performed on a maximum of 360 × 288 pixels (in the case of the NTSC system, 360 × 288 pixels), and the bit rate is a maximum of 15 Mbit / sec.

The type of the input moving image information is a block composed of 8 × 8 pixels to be coded for interlaced images which are interlaced scans and progressive images which are progressive scans. , A macro block composed of 16 × 16 pixels to be detected as a motion vector. When an interlaced image is used as a prediction method, the prediction efficiency is improved by adaptively switching between inter-field prediction and inter-frame prediction. For example, a prediction error is minimized by performing inter-field prediction for a moving image and performing inter-frame prediction for a still image portion.

Further, it is specified that motion estimation is performed by a motion search with half-pixel accuracy. Further, when encoding the prediction error signal, the prediction error signal is transformed by DCT and the transform coefficient is encoded. When the DCT is performed on the frame in macroblock units (frame DCT), The DCT is performed separately for each field (field DCT), and the one in which the amount of encoded information becomes smaller is adaptively selected. This determination may be performed after actually obtaining the respective coded information amounts. However, in general, the amount of calculation is increased by comparing and determining the power of the high frequency component of the difference signal. Is suppressed.

When a picture is reproduced, the structure of a coded picture for enabling random access to decode the picture is a GOP composed of a plurality of consecutive frames.
(Group of Pictures). As a type of an image constituting a GOP, an intra-image encoded image (I image) which is normally encoded only with its own image information, a forward I image or a P image as a reference image, and a forward motion on a time axis. Predicted encoded image (P image) encoded using prediction,
Images classified into three types of bidirectionally predicted coded images (B images) coded using forward and backward motion prediction on the time axis using past and future I or P images as reference images It is.

As an image sequence in a GOP,
For example, when M = 3, the image input order (F0, F1, F2, F3, F4, F5,...) And the image encoding type (I image, P image , B image), F0 (B image), F1 (B image), F
2 (I image), F3 (B image), F4 (B image), F5 (P
), And two frames of the B image are generated continuously between the preceding and succeeding I or P images. Also, M =
In the case of 2, the B image is continuously composed of one frame, M
If = 1, it is composed of only I and P images.
On the other hand, in the case of M> 3, the number of consecutive frames of the B image increases by the numerical value increase.

Here, in this embodiment, the image coding condition is M = 3 as an image sequence represented by the M value.

The encoding process for each frame is performed by 1
Real time processing is realized by staying within the frame period .

Also, the B image at the head of the GOP (F0, F1)
Is to perform prediction by forward prediction and backward prediction using the most recent past I image or P image of the previous GOP as a reference image, or to perform prediction only by backward prediction. In this embodiment, For the sake of simplicity, it is assumed that prediction is performed only by backward prediction.

Next, the operation of this embodiment will be described in detail with reference to FIG. FIG. 2 is a timing chart showing processing of main circuit blocks in FIG.

In FIG. 1, the encoding target is 4:
Input image information 1a expressed in the 2: 2 format or the 4: 2: 0 format is input from the input terminal 1 in the display order. Here, in this embodiment, the maximum value of the M value is limited to 3, for example, when M = 3, the input image information
As described above, the relationship between the frame order of the report 1a and the image sequence is F0 (B image), F1 (B image), F2 (I
Image), F3 (B image), F4 (B image), F5 (P image),...

When the input image information 1a is represented in 4: 2: 2 format by the frame order rearrangement control circuit 2 and the storage medium 3, the input image information 1a is converted into 4: 2: 0 format and 4: 2: If the format is 0, the format is not converted, and is temporarily stored and held in the storage medium 3 in accordance with the instruction 2a of the frame order rearrangement control circuit 2, and the frame or field unit corresponding to the encoding order in accordance with the image sequence according to the M value. And is read out in units of macroblocks by block scan and output as input image information 3a.

Here, in the output frame order of the input image information 3a, since the coding type of the first two frames F0 and F1 is a B image, it is necessary to encode the frame F2 which is an I image first. Therefore, F2 (I image), F0
(B image), F1 (B image), F5 (P image), F3
(B image), F4 (B image),...

In this case, at least three frames are prepared as the capacity of the storage medium 3 (for example, a frame memory) necessary for rearranging the frame order and performing the block scan. Further, the output timing of the input image information 3a is determined so that the encoding process for each frame falls within one frame.

The frame order rearrangement control circuit 2 encodes an I image, a P image, and a B image as two types of bits represented by "01", "10", and "11", respectively, as types of images to be encoded. Image condition information 2b and, as will be described later,
At the time of encoding the B image, a decoding permission signal 2c indicating whether or not to perform local decoding processing is output.

The storage medium control circuit 5 and the recording medium 13
In accordance with the coded image condition information 2b, input image information 3a input in units of macroblocks scan-converts image information corresponding to an I image or a P image, and stores and holds it. At this time, image information 13a is output in a range in which a motion search is performed, using past and future I images or P images as reference images for macroblocks to be encoded.

Here, when encoding a B image, forward and backward motion prediction is performed on the time axis. Therefore, two frames of past and future I images or P images are required. Is stored for at least two frames.

When the motion detection is obtained in multiple stages,
In the initial stage, if the input image stored and held in the recording medium 13 is used as a reference image and the accuracy of the motion search is reduced, the capacity of the storage medium 13 is:
At least as much as the search accuracy is reduced, the capacity can be reduced. For example, as a search precision, if the search is performed roughly in the horizontal direction with two pixel precision, the capacity required for the recording medium 13 can be reduced by half. Further, at the time of motion search, if only the luminance signal is targeted, the capacity for storing and holding the color signal can be reduced.

Next, the type of each image (I image, P image,
The operation of each circuit for each (B image) will be described.

First, for example, the processing of an F2 frame in the case of an I image will be described.

As described above, since intra-picture coding (intra-macroblock) is performed on an I-picture in units of macroblocks, the motion detection circuit 4 and the predicted picture generation circuit 7 perform processing other than processing necessary for picture transfer. The input image 7b is output without any processing. The coded image determination circuit 8 outputs a selection signal 8a for selecting the input image 7b as a coded image because the input image 7b is an intra macro block.

Further, when the input image is an interlaced image, whether the DCT is to be performed in a frame unit or a field unit as a block unit is determined by comparing the power of the high frequency component of the difference signal in macro block units. The decision and block configuration selection signal 8b is output.

In the coded image selection circuit 9, the input image 7b is selected in accordance with the output signal 8a of the coded image determination circuit 8, and the block is selected in units of frames or fields in accordance with the instruction of the block configuration selection signal 8b. Is output as encoded image information 9a.

Here, at the time of encoding an I image, the motion detection circuit 4 and the predicted image generation circuit 7 need to perform motion search, generate a predicted image, and calculate predicted difference image information between the input image and the predicted image. This is to stop the processing.

The coded image information 9a is input to the DCT circuit 10
And by the quantization circuit 11 performs the processing of the quantization of the transform coefficients discrete cosine transform block by block, and outputs the quantization information 11a.

In the variable-length encoding circuit 12, the quantization information 11a is used for encoding prediction conditions, motion vector information,
The variable-length coding process is performed together with coding conditions such as DCT block configuration conditions, and the resultant is output as coded information 12a represented by a bit sequence. The obtained encoded information 1
2a is output from the output terminal 17 as encoded information F2_e for the F2 frame.

On the other hand, the quantization information 11a output from the quantization circuit 11 is supplied to the inverse quantization circuit 1 according to the encoding conditions.
4 and the IDCT circuit 15 perform the reverse processing of the DCT circuit 10 and the quantization circuit 11 to obtain the predicted difference image information 1
It is output as 5a. In the decoding process for the I image,
The decoded image generation circuit 16 calculates the predicted difference image information 15
a itself is the locally decoded image information 16a for the I image.
Output as This locally decoded image information 16a is
As a local decoded image F2_ed for 2 frames, it is stored and held in the recording medium 6 in accordance with an instruction 5b from the storage medium control circuit 5.

Next, the case of a P image will be described. The frame in this case is, for example, the F5 frame in FIG.

In this case, the F2 frame, which is the past I image, is used as a reference image when performing a motion search.

First, the input image information 3a of the F5 frame
Are supplied from the recording medium 3 to the motion detection circuit 4 in macroblock units. On the other hand, this F5 of the input image information 3a
Since the frame is for the P image, the storage medium control circuit 5 causes the F image, which is the I image on the recording medium 13, to be stored.
Scan conversion is performed for an area different from the area stored and held by the two frames, and the area is stored and held.

The motion detection circuit 4 performs motion vector detection for each macroblock by a motion search with half-pixel accuracy in frame and field units. Here, there is no problem if the motion search is performed in the whole area within the frame. However, since the motion search is performed in real time, the search range is limited to the peripheral area of the macroblock or the surrounding macroblock. The time required for the search is reduced by using a motion vector that is a result of the motion search for the prediction vector as a prediction vector and limiting the motion vector to a region around the region indicated by the prediction vector.

Further, the motion search is performed in multiple stages. As the search accuracy and the search range in performing the motion search, in the first stage, the search is performed with a plurality of pixel accuracy, for example, two pixel accuracy, and the first stage motion vector is obtained. In the next stage, the search range is narrowed to the peripheral area indicated by the first-stage motion vector, and the next-stage motion vector is determined by motion search with integer pixel accuracy.
Similarly, in the subsequent stage, the search range is narrowed to the peripheral area indicated by the next-stage motion vector, and the motion vector is determined by a motion search with half-pixel accuracy. Further, it determines the field prediction or the frame prediction as the prediction condition according to the prediction accuracy as the motion search result.

In this embodiment, for the sake of simplicity, the following steps will be omitted from the description.

That is, for each macroblock, image information that matches the search accuracy and the search range in the input image of the F2 frame is read from the recording medium 13 as the first stage processing. In this case, although a coarse search is performed by thinning out reference pixels, a motion vector is detected by performing a search over a wider range within the processing time, and the motion vector is output as a first-stage motion vector 4b.

In the latter stage of the motion search, the motion search is performed with half-pixel accuracy. Therefore, the control signal 5b of the storage medium control circuit 5 controls the local portion of the F2 frame stored in the recording medium 6 in accordance with the initial stage motion vector 4b. among the image information F2_ed as decoded image, the first-stage motion vector 4
At the position indicated by b, image information of 18 × 18 pixels which is larger by a half pixel than the macroblock size is output as the reference image 6a. The motion detection circuit 4 generates a reference image with half-pixel accuracy from the reference image 6a with integer pixel accuracy by interpolating between pixels, and applies the obtained reference image with half-pixel accuracy to MPEG-2. A motion search with half-pixel accuracy, which is the search accuracy of, is performed, and a motion vector and a prediction mode are specified. In this case, the size of the reference image may be further increased to compensate for the motion vector search accuracy in the first stage.

As described above, the image information indicated by the specified motion vector is output from the motion detection circuit 4 as the predicted image 4c, and the image information of the input macroblock is output as the input image 4a. Further, although not shown, the obtained motion vector information with half-pixel accuracy and the prediction mode are output to the variable length encoding circuit 12.

In the predicted image generation circuit 7, the difference between the input image information 4a and the predicted image information 4c is obtained for each macro block, and the predicted difference image is output as information 7a .
Also, input image information 7b and predicted image information 7c are output.

The coded image determination circuit 8 compares the predicted difference image 7a with the input image 7b, and determines the image having a smaller code amount when coded, among them, as the coded image. This is the input image 7
b, an intra macroblock, a prediction coded image 7a, a prediction coded image (non-intra macroblock), and motion prediction, the pixel values of the prediction difference image information 7a are all 0. if it becomes the selection signal 8a coding is unnecessary images (skipped macroblocks) are output. Further, when the input image is an interlaced image, whether the DCT is to be performed on a frame basis or on a field basis as a block configuration is determined by comparing the respective coding amounts on a macroblock basis with a minimum condition. Is determined, and a block configuration selection signal 8b is output.

In the coded image selection circuit 9, the input image 7b or the prediction difference image 7a is selected according to the output signal 8a of the coded image determination circuit 8, and a block is formed according to the instruction of the block configuration selection signal 8b. , Are output as encoded image information 9a. This encoded image information 9a
Is processed by the DCT circuit 10 and the quantization circuit 11 as in the case of the I image, and is output as quantization information 11a.

In the variable-length encoding circuit 12, the quantization information 11a is used for encoding prediction conditions, motion vector information,
The variable-length coding process is performed together with coding conditions such as DCT block configuration conditions, and the resultant is output as coded information 12a represented by a bit sequence. The obtained encoded information 1
2a is coding information F5 for the P image of the F5 frame.
_e from the output terminal 17.

On the other hand, the quantization information 11a is subjected to inverse quantization processing and inverse DCT processing by the inverse quantization circuit 14 and the IDCT circuit 15 in accordance with the encoding conditions in the same manner as the processing for the I-picture, and It is output as difference image information 15a.

In the decoded image generation circuit 16, in the case of a non-intra macro block,
The prediction difference image information 15a and the prediction image information 7c of the block corresponding to the prediction difference image information 15a are added and output as the locally decoded image information 16a. In the case of an intra macroblock, the prediction difference image information 15a is output as locally decoded image information 16a. Further, in the case of a skip macro block, the predicted image information 7c is output as the locally decoded image information 16a. Such local decoded image information 16a obtained in the as locally decoded image F5_ed for F5 frame, under the instruction 5b of the storage medium control circuit 5, stored and held in the recording medium 6
You .

Next, the case of the B image will be described. In this case, the frames in FIG.
There are three frames.

In this case, the encoding of the B image is
It performs forward and backward bidirectional motion prediction on the time axis using a future I image or P image as a reference image. In the case of the F0 frame, as described above, GOP
Since this is the first B image, its prediction direction is only backward prediction. Therefore, the F2 frame, which is a future I image, is used as a reference image when performing a motion search.

The basic operation of predictive encoding is as described above for F
The same processing as in the case of the 5-frame P image is performed. In the case of the F0 frame, the input of the F0 frame is performed.
The image information 3a is sequentially supplied to the motion detection circuit 4. The prediction direction is only backward prediction, and as a reference image,
The F2 I image information stored and held in the recording medium 13, which is the input image of the F2 frame, is
Image information F2_ed respectively used are stored and held in the recording medium 6 where a locally decoded image of a frame, the motion search, the either the motion vector and the frame prediction or field prediction of the half pixel precision prediction mode Identify.

Here, since the processing for the forward prediction becomes unnecessary, the processing for the forward prediction such as the memory access processing for taking in the reference image or the motion search processing may be stopped. In addition, in the case of the leading B image, if the prediction direction is limited to only backward prediction, the motion estimation is performed in the same manner as the prediction process for the P image, so that a wide range motion search can be performed even for the B image. Good.

On the other hand, in the case of the F3 frame, forward prediction and backward prediction are respectively performed. In the case of forward prediction, the F2 frame is used, and in the case of backward prediction, the F5 frame is used as a reference picture. In the case of forward prediction, image information that matches within a search range set by the input image of the F2 frame as the reference image is read from the recording medium 13, a motion vector is determined by motion search, and the motion vector information 4b is output. .

[0082] In a subsequent stage, according to the first stage of motion vector information 4b, reference image information F2_ed that stored and held in the recording medium 6 where a locally decoded image of the F2 frame
It reads out as an image, generates a half-pixel accuracy reference image by interpolating between pixels, performs a motion search on the obtained half-pixel accuracy reference image, and obtains a half-pixel accuracy motion vector and prediction mode for forward prediction. To identify. Similarly, in the case of backward prediction, a motion vector and a prediction mode with half-pixel accuracy are specified by using the F5 frame as a reference image and performing the same processing as in the above-described forward prediction.

Here, backward prediction and forward prediction are completed within the same frame period by making the motion search range of the first stage narrower than that of the P image.

In the predicted image generation circuit 7, in the case of the F0 frame, only the backward predicted image is used as the predicted image, and the input image information 4a and the predicted image information 4c are obtained for each macroblock.
Is obtained, and the prediction difference image 7a, the input image 7b, and the prediction image information 7c are output. In the case of the F3 frame, a difference between the input image information 4a and the predicted image information 6a is obtained for each macroblock.

Here, three types of predicted image information, that is, predicted image information at the time of forward prediction and backward prediction, and bidirectional predicted image information obtained by interpolation from these predicted image information, respectively, are input image data. The difference from the information 4a is calculated, and the sum of the absolute values (hereinafter, referred to as the absolute value difference sum) is obtained as the difference value for each pixel in the macro block.
Among the three types of prediction image information, the one having the smallest absolute value difference sum is determined as the prediction direction and the prediction image information , and the prediction difference image 7a, the input image 7b, and the prediction image information 7c are output.

The coded image determination circuit 8, the coded image selection circuit 9, the DCT circuit 10, the quantization circuit 11, and the variable-length coding circuit 12 perform the same processing as the above-described processing on the P image when the B image is coded. Thus, the encoded information F0_e or the encoded information F3_e for the F0 frame or the F3 frame is generated and output from the output terminal 17.

On the other hand, the inverse quantization circuit 14 and the IDCT circuit 1
5 and the decoded image generation circuit 16 may generate a locally decoded image for the B image by the same processing as that for the P image, but the B image is a reference image for encoding other images. It is not used as an image, and encoding of an input image can be realized even if there is no locally decoded image for the image. Therefore, assuming that the 1-bit decoding permission signal 2c represents "1" when performing the local decoding process and "0" otherwise, "0" is set at the encoding process timing for the B image. In this case, the local decoding processing operation for the B image is stopped. This allows for inverse quantization and IDC
T arithmetic processing, addition processing of the predicted image and the predicted difference image in the decoded image generation circuit 16, local decoded image information 1
The output of the decoded image information 16a to the storage medium 6 by the output of the storage medium control circuit 5 and the storage medium control circuit 5 is stopped. At this time, the output of the predicted image information 7c for the B image to the decoded image generation circuit 16 may be stopped.

The monitor output circuit 18 converts the input moving image information 1a input in the display frame order into the monitor output 1
8a is output from the output terminal 19. When the input moving image information 1a is in 4: 2: 0 format,
You may make it convert into 4: 2: 2 format and output.

As described above, in this embodiment, the MPE
In a moving picture coding apparatus such as G-1 / MPEG-2, all the circuits prepared for moving picture coding processing are not always operated, but the coding is performed according to the coded picture conditions. It is possible to appropriately stop the operation of the circuit block that is not the target of the conversion processing. As a result, even when the encoding device according to the present embodiment is realized by, for example, a one-chip LSI, each circuit block can be efficiently operated at the time of moving image encoding processing, and wasteful consumption is achieved. Power consumption can be easily reduced and power consumption can be reduced.

Although a one-chip LSI can store a memory such as a frame memory in one chip, if the cost of the chip is reduced, the chip size is increased by the memory capacity. It is uneconomical in terms of manufacturing costs. For this reason, the LSI and the external memory are generally realized in a configuration in which the LSI and the external memory are separated. However, according to this embodiment, the unnecessary external memory access is stopped, so that unnecessary data lines and control signal lines are not used. The occurrence of flapping can be reduced, and not only power consumption but also power supply noise and radiation noise can be easily reduced.

In this embodiment, only the case where M = 3 has been described as a GOP structure image sequence.
Even when B = 2, it is needless to say that the same effect can be obtained by performing the same processing as described above and stopping the operation relating to the generation of the locally decoded image when encoding the B image.

In this embodiment, the recording medium 3
The storage capacity for storing and holding frames is provided. Of course, even if M> 3, the storage capacity is increased to increase the number of frames that can be rearranged in the frame order. It goes without saying that the effect of (1) can be obtained.

Further, in this embodiment, DCT is used for the encoding and decoding processes. However, the present invention is not limited to this, and other encoding and decoding methods may be used. Needless to say, low power consumption can be achieved by the above control method.

FIG. 3 is a block diagram showing a second embodiment of the moving picture coding apparatus according to the present invention, a simplified decoded picture monitoring method and a control method using the same, and reference numeral 20 denotes a monitor output control circuit; Is a monitor output selection circuit.
The same reference numerals are given to the portions corresponding to and the duplicate description will be omitted.

FIG. 4 is a timing chart showing the processing in the main circuit blocks in FIG. 3, and the same reference numerals are given to those corresponding to FIG.

In FIG. 3, as in the embodiment shown in FIG. 1, the input image information 1a for three frames input in the display order is stored and held in 4: 2: 0 format on the recording medium 3. 6, the most recent past input image information 1 is obtained by the encoding process in the embodiment shown in FIG.
a of two frames obtained by performing local decoding on the coded information of
The locally decoded image information is stored and held, and the storage media 3 and 6 store the image information in the 4: 2: 0 format in the order of scanning of the images.

The monitor output control circuit 20 controls the reading of the storage media 3 and 6 at a predetermined timing according to a monitor display mode signal 2d, which will be described later, from the frame order rearrangement control circuit 2, and the input image information 3b and the local decoding The converted image information 6b is output in the scanning order and the display frame order.

On the other hand, the monitor output selection circuit 21 selects one of the input image information 3b and the locally decoded image information 6b in the display order according to the monitor output selection signal 20c from the monitor output control circuit 20. Is output as a monitor display output 21a. Here, as in the first embodiment in FIG. 1, the monitor output circuit 18
Image information expressed in 4: 2: 0 format is 4:
The data may be converted into a 2: 2 format and output.

First, the operation of this embodiment in the case of M = 3 as a GOP structure image sequence will be described with reference to FIG.

Here, for the I image and the P image,
The decoded image information 6b (F2_ed, F5_ed,...) For the reference image at the time of encoding, which is stored and held in the recording medium 6, is represented by B
With respect to the images, the input image information 3b (F0, F1, F3, F4,...) Stored and held in the recording medium 3 is output as the monitor display output 21a. The information is read out in the scanning order, and the monitor output selection circuit 21 outputs the input image information 3b and the locally decoded image information 6
b, and monitor display output 21a (F0, F1, F2_ed, F3, F4, F5_
ed,...) (monitor display mode 1 in FIG. 4).

FIG . 5 is a timing chart showing processing in the case where M = 2 in this embodiment.

In the figure, when M = 2, the relationship between the input image frame order and the image sequence is F0 (B image), F1 (I image), F2 (B image), F3 (P ), F4 (B image), F5 (P image),..., And the I image or the P image is generated every two frames. Therefore, in the monitor output selection circuit 21, the input image information 3b (F0, F
2, F4,...) And the locally decoded image information 6b (F1_e
d, F3_ed, F5_ed,...) to switch the monitor display output 21a (F0, F1_ed, F2, F3_ed, F4,
F5_ed,...) (Monitor display mode 1 in FIG. 5 ).

FIG. 6 is a timing chart showing processing in the case of M = 1 in this embodiment.

In the figure, when M = 1, the relationship between the input image frame order and the image sequence is as follows: F0 (I image), F1 (P image), F2 (P image), F2 (P image)
F3 (P image), F4 (P image), F5 (P image), ...
And is composed of an I image or a P image.
In addition, the encoding order matches the display order, and the rearrangement processing of the frame order is not required. Further, the local decoded image information 16
a is generated for all frames. Therefore, in the monitor output selection circuit 21, the locally decoded image information 6b (F0_ed, F1_ed, F2_ed, F3_ed, F4_e) is displayed in the display order.
d, F5_ed,...) are output as the monitor display output 21a (monitor display mode 1 in FIG. 6).

As described above, this embodiment has the same effect as the first embodiment, and also has a new frame memory with only a limited memory capacity, which is a frame memory used for coding moving images. Input information and local decoded image information stored and held in the recording media 3 and 6 can be diverted for monitor display without adding
In addition, it is possible to selectively output the images together with the image display order, and it is possible to easily obtain the monitor display output of the locally decoded image information. Thereby, the monitor display is performed in the display frame order, so that the monitor output can be obtained as a smoother and more natural image of the moving image.

Although the operation status of the encoding process cannot be visually monitored for the B image because the input image information is output on the monitor, the I image or the P image used as the reference image is not encoded. Since the encoding status can be sequentially monitored, the encoding operation status can be visually monitored, and in the unlikely event that a defect occurs in the encoding process of the reference image and a defect propagates to a subsequent encoded image. Can be verified early.

Further, when M = 1, the decoded image information of all frames can be displayed on a monitor and output, and the encoding status of the encoded image can be monitored sequentially, so that the encoded image is visually encoded. The operation status can be monitored.

Here, although not shown in this embodiment, it is needless to say that by increasing the storage capacity of the recording medium 3 and increasing the number of frames that can be rearranged in the frame order, M>
Even in the case of 3, the decoded image information for the I image and the P image is displayed on the monitor by the same processing shown in this embodiment.

The frame memories serving as the storage media 3 and 6 are used for moving picture coding. The access timing for moving picture coding processing read out by block scanning in macroblock units and the display scanning order This control is performed so as not to conflict with the access timing for the monitor display output to be read. However,
Needless to say, by increasing the access speed of the used frame memory, access timing control without conflict can be easily achieved.

Here, FIG. 4 in the second embodiment is shown.
The other monitor display modes in FIGS.

First, in FIG. 4, when M = 3,
Since the I picture or P picture which is locally decoded image information is updated every 3 frames, the local decoding on the local decoded I picture or P picture to the most recent past is stored and held in the recording medium 6 Image information 6b (F2_ed, F
5_ed,...) Are displayed in monitor order in the order of encoding.
And outputs the information of the same frame for three consecutive frames, F2_ed, F2_ed, F2_ed, and F5_e.
d, F5_ed, F5_ed,... (monitor display mode 2 in FIG. 4).

In FIG. 5, when M = 2, the locally decoded image information is updated every two frames. Therefore, the information of the same frame is continuously displayed on the monitor display output 21a (F1_ed, F1_ed, F3_ed, F3_ed, F5_e
d, F5_ed,...) (monitor display mode 2 in FIG. 5).

In FIG. 6, when M = 1, the mode
As in the case of the monitor display mode 1 , the decoded image information 6b (F0_ed, F1_ed, F2_ed, F3_ed, F4
_ed, F5_ed,...) are output as the monitor display output 21a (monitor display mode 2 in FIG. 6) .

As described above, in this embodiment, since the same frame is repeatedly displayed in the display order, the continuity of the moving image is impaired, but the I-frame used as the reference image is lost.
By displaying only the decoded image of the image or the P image on the monitor, it is possible to easily monitor the encoding operation status visually. As a result, even if a failure occurs in the encoding processing of the reference image and the failure propagates to a subsequent encoded image, verification can be performed early.

FIG. 7 is a block diagram showing a first embodiment of a moving picture encoding / reproducing apparatus, a simplified decoded picture monitoring method and a control method using the same according to the present invention. Terminal, 23 is a control signal selection circuit, 24
Is a data selection circuit, 25 is a storage medium in which the storage media 3, 6, and 13 in the previous embodiment are integrated into one and controlled by a common control signal, and 26 is an input of reproduced encoded information. Input terminal 27 is a variable length decoding circuit,
Reference numeral 8 denotes a selection circuit, and portions corresponding to those in FIG.

FIG. 8 is a memory layout diagram of the storage medium 25 in FIG. 7, and FIGS. 9 to 11 are timing diagrams showing processing of main circuit blocks in FIG.

The first video encoding / reproducing apparatus has
Is provided with a decoding mode for inputting and decoding reproduced encoded information, and hereinafter, the maximum possible M value is M = 3.

In FIG. 7, the storage medium 25 is such that the storage media 3, 6, and 13 are integrated into one and controlled by a common control signal. For example, an address arrangement as shown in FIG. Stipulated.

Here, the PAL system (CC of 625 lines)
In the case of supporting up to 4: 2: 0 format of IR 601 standard, 720 luminance signals Y are used per frame.
× 576 pixels, and each of the color difference signals Cb and Cr is 360 × 288
It is a pixel and uses a capacity of 8 bits for each pixel.
Also, at the time of encoding, as a reference image at the first stage of motion detection,
As described above, since the luminance signal in the scanning direction is thinned out, the capacity becomes 360 × 576 pixels. Therefore, as an arrangement at the time of encoding, as shown in FIG.
Y2, C2, Y3, C3, Y4, C4 areas as storage media 6, Y5, C5, Y6, C6 areas as storage medium 6, storage medium 1
3, the Y0 and Y1 areas are allocated. As shown in FIG. 8B, the arrangement of the storage medium 6 at the time of decoding will be described later, but the storage mediums 3 and 13 are not used, and the newest used as a reference image as decoded image information. Y6, C6, Y5, and C that store and hold decoded image information of an I image or a P image that has been decoded in the past.
Y for storing and holding decoded image information of 5 areas and B image
7, C7, Y8 and C8 areas are allocated.

A processing mode instruction signal 22 for instructing from the input terminal 22 the encoding mode or the decoding mode as the processing mode.
a is input, and in response, the frame order rearrangement control circuit 2 outputs a mode signal 2e indicating whether it is an encoding process or a decoding operation. In this case, the mode signal 2e is, for example, "1" for an encoding processing operation, and "0" otherwise.

In the control signal selection circuit 23 , FIG.
3 , a control signal 5a for input reference image information, a control signal 5a for locally decoded image information,
b and the input image information readout control signal 20a and the locally decoded image information readout control signal 20b which are monitor output control signals. When each control signal is input, it is converted into the corresponding area address, and the address and write are performed. It outputs a read control signal 23a and a data selection control signal 23b, respectively.

In the data selection circuit 24, when writing to the storage medium 25, the data selection control determined by the frame order rearrangement control signal 2a, the input reference image information control signal 5a, and the locally decoded image information control signal 5b. Signal 23b
According to the above, one of the input image information 1a, the input image information 3a of the reference image , and the locally decoded image information 16a, which are the data to be written to the storage media 3, 6, and 13, is selected and output as the write data 24a.

At the time of reading from the storage medium 25, the read data 25a includes a frame order rearrangement control signal 2a, an input reference image information control signal 5a, a locally decoded image information control signal 5b, and an input image information read control. Signal 20a, local input image information 3 according to data selection control signal 23b determined by decoded image information read control signal 20b.
a, the locally decoded image information 6a and 6b, the input reference image information 13a , and the input reference image 3b .

At the time of encoding when M = 3, the selection circuit 28 outputs the quantization information 11a to the selection output 28b (1
Although the operation of 1a) is added, as shown in FIG. 10, video encoding and monitor output are performed by the same processing as shown in the first and second embodiments of the video encoding device. .

On the other hand, at the time of encoding when M = 2, as shown in the first and second embodiments of the moving picture encoding apparatus, the capacity of the storage medium 3 necessary for the frame order rearrangement processing is increased. For two frames, the storage medium 25 is arranged in the Y2, C2, Y3, and C3 areas in the storage medium 3, the Y4, C4, Y5, C5, Y6, and C6 areas in the storage medium 6 and the storage medium 13 in the storage medium 13. Each of the Y0 and Y1 areas is assigned. As a result, as shown in FIG. 11, the local decoding process is also performed on the B image so as to be stored and held in the Y4 and C4 areas of the storage medium 6, and the local decoded image information read control signal from the monitor output control circuit 20 is output. Y4 ~ 6, C4 ~ 6 depending on 20b
By reading the locally decoded image information of the area, the monitor display output 21 the locally decoded image information for all images a
Can be obtained as

Next, the decoding operation of this embodiment having the above configuration will be described. Here, the case where M = 3 is set, A case where decoding processing is performed on encoded information that has been encoded under the same encoding conditions as the encoding conditions described in the embodiment will be described with reference to FIG.

[0127] First, F2_e inputted from the input terminal 26 the encoding frame order (I picture), F0_e (B picture), F
1_e (B picture), F5_e (P picture), F3_e (B picture), F4_e (B picture), the reproduction coded information 26a is represented by bit sequence entered in ......, the variable length decoding circuit 27 are processed decoded, the coding conditions in macro block units, the predicted conditions and the half-pixel precision motion vector information 27a, such as quantization information 27b is output being played. The quantization information 27b is output by the selection circuit 28
It is output as a selection output 28b (27b) .

[0128] Further, the reproduced quantization information 27b is by the inverse quantization circuit 14 and the IDCT circuit 15, by the inverse quantization and IDCT processing on each macro-block is performed, the decoded differential image data 15a Is generated.

In the case of the decoding process of the F2_e frame which is an I image, since the decoding process is an encoding process within a frame and all macroblocks are decoded, the decoded image generation circuit 16 So it outputs the decoded differential image data 15a itself as the decoded image information 16a. Et al is, according to a local decoded image information control signal 5b, the control signal selection circuit 23, decoded image data F2_ed to memory retention area Y6, address corresponding to the C6 of F2_e frames in the storage medium 25 is generated Is output as a signal 23a together with the write control signal. Also, the decoded image information 16 a
b, the selected output 24a is output by the data selection circuit 24.
And the address and write control signal 23 in the storage medium 25.
is stored according to a.

In the case of the decoding process of F5_e which is a P image, F2 which is the decoded image information of the past I image
The _ed frame is used as a reference image, and the storage medium control circuit 5, the control signal selection circuit 23, and the storage medium 25 are used for each macroblock in accordance with the prediction condition and the instruction of the motion detection circuit 4 based on the motion vector information 27a.
The data selection circuit 24 reads out the image information of the 18 × 18 pixel area as the reference image 6a because the image information of the F2_ed frame is the area indicated by the motion vector information and has half-pixel accuracy. Then, the motion detection circuit 4 generates image information with half-pixel accuracy, and outputs 16 × 16 pixel image information 4c indicated by the motion vector via the predicted image generation circuit 7 as predicted image information 7c.

In the decoded image generating circuit 16, the prediction image information 7c for the skip macro block, and the decoded difference image information 15a for the intra macro block, according to the coding condition 27b for each macro block.
In the case of a non-intra macro block, the result of addition of the predicted image information 7c and the decoded difference image information 15a is output as decoded image information 16a, respectively, and the storage medium is processed in the same manner as in the case of the I image. In the Y5 and C5 areas, decoded image information F5_ed of the F5 frame is stored and held.

In the case of the F8_e frame which is a P image, F5_ed which is decoded image information of a past P image
The decoded image information 16a is generated by the same process as the decoding process of the F5_e frame using the frame as the reference image, and is stored and held in the Y6 and C6 areas of the storage medium 25 as the decoded image information F8_ed of the F8 frame. .

In the case of decoding the F0_e and F1_e frames, which are B images, only the backward prediction is used as the prediction direction. Therefore, the F2_ed frame, which is the decoded image information of the future I image, is used as the reference image. The decoded image information 16a is generated by the same processing as the decoding processing of the P image,
8, Decoded image information F0_ed of F0 frame in C8 area
, The decoded image information F1_ed of the F1 frame is stored and held in the Y7 and C7 areas, respectively.

[0134] On the other hand, a B picture F3_e, when the decoding of F4_e frame, since corresponding to the bidirectional prediction, and F5_ed frame is past F2 _Ed frame and future P picture is an I picture Is used as a reference image, and encoding conditions, prediction conditions, and motion vector information 27 with half-pixel accuracy in reproduced macroblock units are used.
According to a, a predicted image 4c in each prediction direction is generated as in the case of the P image. The prediction image generation circuit 7, according to the prediction direction 27 a and the predicted conditions, or predicted image information of the forward prediction or the prediction image information of the backward prediction, or determined by interpolation from the forward prediction and backward prediction respectively predicted image Any of the three types of bidirectional predicted image information is selected and output as the predicted image information 7c. As a result, the decoded image information 16a is generated by the same processing as that for the P image, the decoded image information F3_ed of the F3 frame is set in the Y8 and C8 areas of the storage medium 25, and the F4 is set in the Y7 and C7 areas.
Each is stored and held as decoded image information F4_ed of the frame.

Further, the monitor output control circuit 20
The decoded image information stored and held in the storage medium 25 is displayed in the image display order (F0_ed, F1_ed, F2_ed, F3_ed, F
4_ed, F5_ed,...) And read out as image information 6b, which is selected by the monitor output selection circuit 21 and output from the output terminal 19 as reproduced image information 21a.

At the time of decoding a moving image in this embodiment, the frame order rearranging control circuit 2 sends a mode signal 2e indicating whether the operation is an encoding process or a decoding operation.
To the motion detection circuit 4, the predicted image generation circuit 7, the encoded image determination circuit 8, the encoded image selection circuit 9, the DCT circuit 10 ,
The data is output to the quantization circuit 11 and the variable length coding circuit 12.
In accordance with the instruction of the mode signal 2e, at the time of the decoding process, the motion detection circuit 4 which is a circuit block not related to the decoding process.
, A difference image generation process in the predicted image generation circuit 7, a coded image determination circuit 8, a coded image selection circuit 9, a DCT circuit 10, a quantization circuit 11, a variable length coding circuit 12, and an I image or P The interpolation prediction image generation processing in the prediction image generation circuit 7 at the time of image decoding processing, the dequantization circuit 14, the IDCT circuit 15 at the time of a skip macroblock, and the addition processing at the decoded image generation circuit 16 at the time of an intra macroblock are respectively performed. Stop.

As described above, in the embodiment of the moving picture coding / reproducing apparatus, the first and second moving picture coding apparatuses are used.
In addition to the effects obtained in the embodiment, the circuit block used for the moving image encoding process can be effectively used as the moving image decoding circuit block, and it is easy to suppress an increase in the circuit scale. Can be achieved. In addition, during the decoding process, by stopping the circuit blocks that are not involved in the decoding process, it is possible to suppress unnecessary power consumption.

Further, it is possible to share the address and the data without independently controlling each of the plurality of storage media. As a result, when a one-chip LSI is used, the frame memory is changed to the LS from the viewpoint of manufacturing cost.
Even in the case of a configuration arranged outside the I, address lines, data lines, and control signal lines can be shared, and an increase in the number of LSI input / output terminals can be suppressed.

Further, by sharing the address lines, the data lines, and the control signal lines in this manner, for example, a memory area which is not used at all depending on the operating conditions without depending on the assigned area is generated. However, by dynamically changing the control address, it can be easily used for other processing.

Here, in this embodiment, only the operation in the case where the maximum M value is M = 3 has been described. However, by increasing the capacity of the storage medium 25, the coding of the moving image when M> 3 is performed. -It goes without saying that the same effect can be obtained in the decoding process by the same process as described above. Further, only the case where a plurality of frame memories for storing and holding image information are shared as the storage medium 25 has been described, but it is needless to say that a buffer memory used for other purposes can be shared. Needless to say.
Further, there is no need to limit the frame memory capacity per frame as described above.

FIG. 12 is a block diagram showing a second embodiment of a moving picture encoding / reproducing apparatus, a simplified decoded picture monitoring method and a control method using the same according to the present invention.
The same reference numerals are given to the portions corresponding to and the duplicate description will be omitted.

This embodiment differs from the embodiment shown in FIG. 7 in that input image information 1a is added as an input to the monitor output selection circuit 21.

In FIG. 12, when encoding a moving image,
The monitor output selection circuit 21 converts the input moving image information 1a input in the display frame order into the monitor display output 21 as it is.
a. Other operations of the moving picture coding processing and the decoding processing are the same as those of the first embodiment of the moving picture coding / reproducing apparatus shown in FIG.

In the second embodiment, in addition to the effect obtained in the third embodiment shown in FIG. 7, when encoding a moving image, the encoding processing status cannot be confirmed on a monitor display, but is not stored. Since the input image information 3b and the decoded image information 6b stored and held in the medium 25 are not read out, access to the storage medium 25 can be reduced. As a result, the read area of the reference image required for the encoding process can be enlarged, and a wider range of motion search can be easily performed.

[0145]

As described above, according to the present invention,
2. Description of the Related Art In a moving picture coding apparatus such as MPEG-1 / MPEG-2, not all circuit blocks used for moving picture coding processing are constantly operated, but decoding circuits not involved in coding processing are used. The operation of the circuit block can be stopped as needed. As a result, even when this embodiment is realized by, for example, a one-chip LSI, each circuit block can be efficiently operated at the time of moving image encoding processing, and is wasted. Low power consumption can be easily achieved by eliminating power,
Even in the case where the present invention is applied to a storage-cell-type portable video encoding device or the like, the power consumption can be reduced and the usable time can be prolonged.

Further, according to the present invention, a frame memory used for encoding a moving image can be generated in the encoding process without adding any new frame memory with a limited memory capacity. The input image information and the locally decoded image information in the frame memory to be used can be used for monitor display, and a monitor display output of the locally decoded image information can be easily obtained. As a result, since the monitor display is performed in the display frame order,
A monitor output is obtained as a smoother and more natural image of the moving image, and furthermore, since the encoding status for the I image or P image used as the reference image can be sequentially monitored, the encoding operation status can be visually monitored, Even if a failure occurs in the encoding processing of the reference image and the failure propagates to a subsequent encoded image, verification can be performed early.

Further, according to the present invention, when M = 1 as an image sequence, it becomes possible to display locally decoded image information of all frames on a monitor display and to output the encoding status for the encoded image sequentially. Since monitoring is possible, it becomes possible to visually monitor the encoding operation status.

Further, according to the present invention, it is possible to effectively use a circuit block used for moving image encoding as a moving image decoding circuit block, and to suppress an increase in circuit size. Easy to achieve. Further, at the time of the decoding process, it is possible to stop the circuit blocks that are not involved in the decoding process, and it is possible to suppress unnecessary power consumption.

Further, according to the present invention, it is possible to share addresses and data without individually controlling a plurality of storage media. As a result, even when the frame memory is arranged outside the LSI from the viewpoint of manufacturing cost when implementing the one-chip LSI, the address line, the data line, and the control signal line can be shared, and the LSI can be integrated. It is possible to suppress an increase in the number of input / output terminals.

[Brief description of the drawings]

Moving image encoding apparatus that by the invention; FIG and simple decoding image monitor method using the same, a block diagram showing a first embodiment of the control method.

FIG. 2 is a timing chart showing a processing operation in main circuit blocks in FIG.

[Figure 3] moving image encoding device that by the present invention and the simple decoding image monitor method using the same, a block diagram showing a second embodiment of the control method.

FIG. 4 is a timing chart showing a processing operation of the main circuit block in FIG. 3 when M = 3.

FIG. 5 is a timing chart showing a processing operation when M = 2 of a main circuit block in FIG. 3;

6 is a timing chart showing a processing operation of the main circuit block in FIG. 3 when M = 1.

FIG. 7 is a block diagram showing a first embodiment of a moving picture encoding / reproducing apparatus, a simplified decoded picture monitoring method and a control method using the same according to the present invention.

FIG. 8 is a layout diagram showing a memory area of a storage medium in FIG. 7;

9 is a timing chart showing a decoding processing operation when M = 3 of main circuit blocks in FIG. 7;

10 is a timing chart showing an encoding processing operation when M = 3 of main circuit blocks in FIG. 7;

11 is a timing chart showing an encoding processing operation of the main circuit block in FIG. 7 when M = 2 .

FIG. 12 is a second embodiment of a video encoding / reproducing apparatus according to the present invention, a simplified decoded image monitoring method using the same, and a control method.
It is a block diagram showing an embodiment.

FIG. 13 is a block diagram illustrating an example of a conventional video encoding device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Image information input terminal 2 Frame order rearrangement control circuit 3 Storage medium 4 Motion detection circuit 5 Storage medium control circuit 6 Storage medium 7 Predicted image generation circuit 8 Encoded image determination circuit 9 Encoded image selection circuit 10 DCT circuit 11 Quantum Circuit 12 variable length coding circuit 13 storage medium 14 inverse quantization circuit 15 IDCT circuit 16 decoded image generation circuit 17 output terminal of coded information 18 monitor output circuit 19 monitor output terminal 20 monitor output control circuit 21 monitor output selection circuit 22 Input terminal of processing mode 23 Control signal selection circuit 24 Data selection circuit 25 Storage medium 26 Input terminal of encoded information 27 Variable length decoding circuit 28 Selection circuit

Continued on the front page (72) Inventor Koichi Terui 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Hitachi, Ltd. Multimedia Systems Development Division (72) Inventor Yukio Isobe 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Co., Ltd. Hitachi, Ltd. Multimedia System Development Division (72) Inventor Takashi Nishimura 292, Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Hitachi, Ltd. Multimedia System Development Division (56) References JP-A-6-98314 (JP, A JP-A-6-189294 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N ^7/ 24-7/68 H03M ^7/ 30-7/40

Claims (6)

(57) [Claims] 1. An intra-coded image in which input moving image information is encoded using only its own image information for each image frame or field, and a previously encoded image as a reference image on a time axis. Use forward and backward motion prediction on the time axis using forward prediction coded images that are coded using forward motion prediction and past and future intra-coded images or forward prediction coded images as reference images the encoded image conditions a bidirectional predictive coded picture to be coded and, a moving picture encoding apparatus for encoding, the predicted differential image input dynamics image information or the input moving image information and the predicted image information to encode information, and encoding means for generating encoded image information, and the local decoding means for reproducing the local decoded image information, a first storage medium for storing and holding said input dynamics image information, 該局unit decodes Chemical image information Storage medium for storing information
And the image information stored in the first storage medium and the second
Either of the image information stored in the storage medium
Monitor output selection means for selecting and outputting information, reading and controlling the first and second storage media, and
The coded image condition is an intra-coded image or a forward prediction code.
If the image is an encoded image, the image is stored and held in the second storage medium.
Selected and output the locally decoded image information,
If the condition is a bidirectional predictive coded image, the first storage
Select and output the input video information stored and held in the medium.
Monitor output control for controlling the monitor output selecting means.
Video encoding apparatus characterized by comprising a means. 2. An intra-coded image in which input moving image information is encoded using only its own image information for each image frame or field, and a previously encoded image as a reference image on a time axis. Use forward and backward motion prediction on the time axis using forward prediction coded images that are coded using forward motion prediction and past and future intra-coded images or forward prediction coded images as reference images A moving picture coding apparatus for performing coding according to a coded picture condition of a bidirectional predictive coded picture to be coded , comprising input moving picture information or input moving picture information and predicted picture information.
Of predicted difference image information to generate encoded image information
Encoding means, a local decoding means for reproducing the locally decoded image information, a storage medium for storing and holding the locally decoded image information, and a module for selectively outputting the image information stored and held in the storage medium.
A memory output selecting means for reading and controlling the storage medium;
The case is an intra-coded image or a forward prediction coded image
In the case, the locally decoded image stored and held in the storage medium
Information is selected and output, and the coded image condition is a bidirectional prediction code.
In the case of a decoded image, the local decoded image selected and output immediately before
The monitor output selection means is controlled so that the information is output again.
Video encoding apparatus characterized by comprising a monitor output control unit Gosuru. 3. The method according to claim 1 , wherein the input moving image information is stored in each image frame or each image frame.
Encodes only its own image information for each field.
Reference to intra-coded images and previously coded images
Code using forward motion prediction on the time axis as an image
Forward predictive encoded image and past and future intra code
Coded image or forward prediction coded image as a reference image
Encode using forward and backward motion estimation on axis
Depending on the coded image conditions with the bidirectional predictive coded image,
A moving picture coding apparatus for performing coding, comprising: input moving picture information or input moving picture information and predicted picture information.
Of predicted difference image information to generate encoded image information
Encoding means, local decoding means for reproducing locally decoded image information, and display frame information for input moving image information to be encoded by the encoding means.
; And a monitor output means for outputting at arm order
Moving image encoding apparatus. 4. The method according to claim 1 , wherein the input moving image information is stored in each image frame or each image frame.
Encodes only its own image information for each field.
Reference to intra-coded images and previously coded images
Code using forward motion prediction on the time axis as an image
Forward predictive encoded image and past and future intra code
Of images or time forward predictive coded picture as see images
Encode using forward and backward motion estimation on axis
Depending on the coded image conditions with the bidirectional predictive coded image,
Encoding can be performed, and
Coded image, forward predictive coded image or bidirectional predictive coded
Performing playback processing to decode and play back any of the images
A video encoding / reproducing apparatus capable of input video information or input video information during encoding processing.
And encoding of prediction difference image information between image and prediction image information
Generates image information and performs encoding processing during playback processing.
Encoding means for stopping, a local decoding means for reproducing the local decoded image information, a first storage medium for storing and holding said input dynamics image information, a second for storing and holding the該局portion decoded image information Storage media
And image information stored and held in the first storage medium and the second
Any one of the image information stored in the storage medium
Monitor output selection means for selecting and outputting information; reading and controlling the first and second storage media;
The coded image condition is an intra-coded image or a forward prediction code.
If the image is an encoded image, the image is stored and held in the second storage medium.
Selected and output the locally decoded image information,
If the condition is a bidirectional predictive coded image, the first storage
Select and output the input video information stored and held in the medium.
Monitor output control for controlling the monitor output selecting means.
Means, during reproduction processing, the intra-coded image, the forward prediction coded image,
Image or bi-directionally coded image
Video encoding / reproducing apparatus characterized by comprising a decoding reproducing means for raw. 5. The method according to claim 1 , wherein the input moving image information is stored in each image frame or each image frame.
Encodes only its own image information for each field.
Reference to intra-coded images and previously coded images
Code using forward motion prediction on the time axis as an image
Forward predictive encoded image and past and future intra code
Coded image or forward prediction coded image as a reference image
Encode using forward and backward motion estimation on axis
Depending on the coded image conditions with the bidirectional predictive coded image,
Encoding can be performed, and
Coded image, forward predictive coded image or bidirectional predictive coded
Performing playback processing to decode and play back any of the images
Possible a moving picture encoding / playback apparatus, when the encoding process, the input video information or input video information
And encoding of prediction difference image information between image and prediction image information
Generates image information and performs encoding processing during playback processing.
Encoding means for stopping, local decoding means for reproducing locally decoded image information, a storage medium for storing and holding the locally decoded image information, and a mode for selecting and outputting the image information stored and held in the storage medium.
A memory output selecting means for reading and controlling the storage medium;
The case is an intra-coded image or a forward prediction coded image
In the case, the locally decoded image stored and held in the storage medium
Information is selected and output, and the coded image condition is a bidirectional prediction code.
In the case of a decoded image, the local decoded image selected and output immediately before
The monitor output selection means is controlled so that the information is output again.
Monitor output control means for controlling the intra-coded image, the forward predicted coded image during the reproduction process,
Image or bi-directionally coded image
Moving image characterized by comprising decoding / playback means for generating
Image encoding / reproduction device. 6. The method according to claim 1 , wherein the input moving image information is
Encodes only its own image information for each field.
Reference to intra-coded images and previously coded images
Code using forward motion prediction on the time axis as an image
Forward predictive encoded image and past and future intra code
Coded image or forward prediction coded image as a reference image
Encode using forward and backward motion estimation on axis
Depending on the coded image conditions with the bidirectional predictive coded image,
Encoding can be performed, and
Coded image, forward predictive coded image or bidirectional predictive coded
Performing playback processing to decode and play back any of the images
A video encoding / reproducing apparatus capable of input video information or input video information during encoding processing.
And encoding of prediction difference image information between image and prediction image information
Generates image information and performs encoding processing during playback processing.
Encoding means for stopping, local decoding means for reproducing locally decoded image information, and display frame for displaying input moving image information to be encoded by the encoding means.
A monitor output means for outputting the images in the order of the images;
Image or bi-directionally coded image
Moving image characterized by comprising decoding / playback means for generating
Image encoding / reproduction device.