KR100511013B1 - Method and apparatus for recording and playing video images - Google Patents

Abstract

Disclosed is a method for reproducing a plurality of video frames coded according to the MPEG format, in which a recording medium on which recordings are installed is reproduced at a speed different from the nominal reproduction speed, transmitted to a display device, and generated by a reading process. The number is compensated so that the average number per unit time is equal to a predetermined nominal value by repeating or skipping a plurality of B frames.

Description

Method and apparatus for recording and playing video images

The present invention relates to recording and reproduction of image information (video). In particular, the present invention relates to a general problem associated with reproducing image information recorded at a speed other than the normal speed.

As is known, video images (frames) are reproduced on a standard image display device such as a television receiver according to a predetermined format, wherein a plurality of pixels of the frames are reproduced sequentially rather than simultaneously. In particular, a plurality of adjacent pixels are successively recorded to record horizontal lines, and the horizontal lines of the frame are successively recorded (interlaced). Therefore, the time required to reproduce one frame, that is, the time required to record all the image lines constituting the frame, is 1/25 s in the PAL format. As a result, a scene with an actual playback time of 100s corresponds to a sequence of 2500 consecutive frames.

For example, in order to record a video image on a recording medium such as a magnetic tape, an analog method is known in which a plurality of consecutive pixels are recorded adjacent to each other on a recording track inclined with respect to the reproduction direction of the recording medium. According to this, each recording track corresponds to one picture line. With this type of recording, the possibility for slow-motion or fast-motion reproduction is limited. For example, if one wants to reproduce a scene at five times the normal speed, the record carrier passes through the readhead at five times the normal speed. This means that five consecutive images pass through the readhead at the time required for reproducing a single image by the image display device. Accordingly, the images reproduced on the display screen have the form of five fragments of these five successive images, which are displayed on top of each other and separated from each other by distorted image portions.

Recently, a digital recording method called MPEG has been developed. In this method, information associated with an image is digitally coded and compressed and stored linearly on a record carrier, which means that a plurality of consecutive bits are arranged in line with each other in the recording and reading direction of the record carrier. .

When such a recording is reproduced, the plurality of bits to be read does not correspond to the plurality of pixels of the frame to be continuously reproduced. In order to decode and reconstruct picture content consisting of one or more consecutive frames, a plurality of bits must be collected. The buffer memory and the computing device required for this purpose may be installed in the playback device or the display device. At this time, it should be noted that on average, the number of read bits corresponds to 25 frames per second at the normal playback speed of the recording medium. In principle, the recording medium can move faster or slower to pass through the readhead. However, a standard display device such as a television receiver has a problem in that the received signal cannot be converted into a regular frame.

It is, after all, an object of the present invention to provide a method and apparatus for allowing video recordings produced according to the MPEG method to be reproduced at an adjustable reproduction speed within a sufficient margin on a standard display device such as a television receiver.

In particular, it is an object of the present invention to allow a user of such a playback device to change the playback speed almost continuously between slow motion and fast motion playback.

These and other features, features and advantages of the present invention will become apparent from the following description of the preferred embodiments of the method and apparatus according to the present invention given with reference to the following accompanying drawings. will be:

1A to 1E show the amount of information saved by MPEC coding,

2A shows an example of an MPEG sequence,

FIG. 2B shows an example of a sequence for slow motion reproduction derived from the sequence shown in FIG. 2A,

FIG. 2C shows an example of a sequence for fast motion reproduction derived from the sequence shown in FIG. 2A,

3 is a schematic block diagram of an apparatus according to the invention.

Since MPEG coding is already known, only some essential aspects of the MPEG recording process will be briefly described with reference to FIG. In particular, three different types of MPEG coding for one frame will be briefly described. This type of coding is called I coding, P coding, and B coding. in need.

Fig. 1A shows a sequence consisting of a plurality of consecutive frames to be recorded on a recording medium such as a magnetic tape, each frame having sequence numbers N, N + 1, N + 2, and the like. Each frame in the PAL format corresponds to a 1 / 25s playback time.

Fig. 1B schematically shows a magnetic tape 10 in which continuous recording areas R, R + 1, R + 2, etc. are displayed thereon, these areas having the same length, that is, these recording areas R, R + 1, Each of R + 2 and the like includes the same number of bits. In the case where each frame N, N + 1, N + 2, etc. is individually recorded in a digitally coded form, the recording of the sequence shown in FIG. 1A has the form schematically shown in FIG. 1B.

In order to reduce the amount of information displayed in bits / second during the reading process of the digital information and during the transmission of the thus read information to the reproduction apparatus, the image information is compressed before recording. In Fig. 1C, the recording medium 10 is a continuous recording area C, C + 1, C + 2 each having a length less than the length of the corresponding recording areas R, R + 1, R + 2, etc., shown in Fig. 1B. This is shown schematically with the back. In particular, FIG. 1C relates to a situation that occurs when information of each individual frame N, N + 1, N + 2, etc. is compressed into a compressed frame before recording. In such a case, the associated original frames N, N + 1, N + 2, etc. may be derived from the respective compressed frames C, C + 1, C + 2, etc. by decompression and decoding processes. Coded and compressed frames having this feature are called intrinsic frames or I frames.

In general, it is possible to further reduce the amount of information from the fact that a plurality of consecutive frames of a sequence have very large similarities with each other. If one frame only records the difference from its previous frame, less information is required to record such a frame. Such a frame is called a predicted frame or a P frame and includes motion information of a frame portion. 1D schematically shows a state in which frame N is recorded as an I frame and the next two frames N + 1 and N + 2 are recorded as P frames. At this time, for example, the frame N + 1 cannot be reconstructed by decompressing and decoding only the recorded information of the P frame corresponding to the frame N + 1, and for this purpose, the recorded frame of the P frame corresponding to the frame N + 1 cannot be reconstructed. It is clear that it is necessary to synthesize the information with the I frame corresponding to the frame N. However, it is also clear that only a relatively small number of bits are needed to write two consecutive frames N and N + 1.

MPEG further includes a third type of coded and compressed frame called a bidirectional frame or B frame. As schematically shown in Fig. 1E, fewer bits are required to write a B frame than to write a P frame. The original corresponding to this B frame is information of three coded and compressed frames, namely the above-mentioned B frame, the last I frame or P frame preceding it, and the first I frame or P frame following the above B frame. Is reconstructed from.

At this time, it should be noted that the amount of information to be recorded depends on the image content of a plurality of consecutive frames. For example, about 540 Gbits (67.5 Gbytes) are required for reproduction of a one-hour image sequence, but such information amount can be compressed to about 10.8 Gbits (1.35 Gbytes) by the MPEG recording process.

Hereinafter, the reproduction process of the coded video recording will be described with reference to FIG.

2A schematically illustrates an example of a sequence consisting of a plurality of stored video frames. These frames are numbered 1, 2, 3,... As indicated at the top of FIG. 2A. The shape of each frame is shown in the lower part of Figure 2a. For simplicity, the I, P, and B frames are shown as having the same size. The coding structure IBBPBBPBB can be repetitive, in which case a group of nine consecutive frames is called a "Group of Pictures" or GOP. At this time, it should be noted that the length of the GOP (that is, nine frames in this case) in the MPEG is variable. Moreover, the structure of the GOP, in particular the number of subsequent B frames, is variable and depends on the picture content of the frame to be recorded. In addition, while it is determined during recording of a sequence by a recording apparatus (video recorder) which frame should be recorded as an I frame, a P frame or a B frame, it should be noted that the way in which this determination is made is not relevant to the present invention. do. Within the scope of the present invention, it can be assumed that the recorded sequence includes I frames, P frames and B frames with some regular coding structure.

When the thus recorded sequence is reproduced in a normal manner (i.e., in real time) by using a display device such as a television receiver, a plurality of coded frames are continuously read by the playback device (video recorder), and coded. It is supplied to the display device in the form. In addition, each coded frame includes information related to a coding type. As noted above, the recording of an I frame contains more information than the P and B frames, so reading and transmitting the P frame to the display device requires less time than reading and sending the I frame, Reading and transmitting B frames takes even shorter time. However, on average, there are 25 PAL-format frames (both I, P and B frames) that are read and transmitted.

In the display device, after a plurality of received frames are decompressed and decoded, the decoded frames are stored in a buffer memory. The stored plural frames are read out from this buffer memory and displayed, and reproduction of each frame in the PAL format takes 1/25 seconds regardless of whether it is coded into I frames, P frames or B frames.

If it is desired to watch a sequence recorded at a speed other than normal speed (fast motion playback or slow motion playback), it is inappropriate to simply move the recording medium past the readhead at a higher or lower speed. In practice, in such a case, the number of frames stored in the buffer memory may be larger or smaller than the nominal number, while in a standard display device, a plurality of frames are provided at the nominal speed (25 per second in PAL format). Is read from In the present invention, as described later, even when the recording medium passes through the readhead at a higher or lower speed (fast motion reproduction or slow motion reproduction), the average number of coded plural frames given to the display device is nominal. It provides a method to be kept the same as.

3 schematically shows a reproduction apparatus 100 according to the present invention, which is configured to read information recorded on the recording medium 10. The reproducing apparatus 100 includes a means 20 for moving the read head 30 and the recording medium 10 past the read head 30. In the case where the recording medium 10 is a magnetic tape, the conveying means 0 is provided with a motor for driving a winding reel to the tape. In the case where the recording medium 10 is a magnetic or optical disk, the conveying means 20 comprises a motor for rotating the disk.

The playback device 100 includes a control device 40, which may comprise, for example, a microprocessor (hardware oriented or software oriented) suitably programmed. The control device 40 is connected to the conveying means 20 to control the speed when the conveying means 20 moves through the read head 30 to move the recording medium 20.

Moreover, the control device 40 is connected to the readhead 30 to receive information read by the readhead 30, which corresponds to the coded information for the video frame as described above. The control device 40 is further connected to the output 101 of the playback device 100 to selectively apply the information read by the read head 30 to the output 101, as will be described later.

The playback device 100 includes a playback speed selection switch 50 connected to the control device 40, which can be driven by a user. The selection switch 50 may have various configurations. In one configuration, the selector switch 50 can be configured as a pointer that can move along the scale, which drives a wiper that can move, for example, along a resistance track. In another configuration, the selection switch 50 may be an alphanumeric keyboard. Moreover, these structures can also be combined, as will be apparent to those skilled in the art to which the present invention pertains.

Using this selection switch 50, the user can command the controller 40 to select the desired playback speed. Normally, the user will select a normal playback speed, indicated by " N " in FIG. The control device 40 controls the transfer means 20 such that the recording medium 10 moves at a substantially constant predetermined speed, referred to below as the normal speed. The video signal received from readhead 30 is thus transmitted in its entirety to output 10.

The control device 40 is also configured to monitor the data stream received from the readhead 30 and to measure the average number of frames per unit time in this data stream. The controller 40 compares the measured average value with a predetermined fixed value, hereafter referred to as a nominal value, wherein the nominal value has a value of 25 per second in the PAL format. If the measured mean value is greater than this nominal value, the control device 40 controls the conveying means 20 to cause the recording medium to move past the readhead 30 more slowly, and the measured mean value is greater than this nominal value. In the small case, the control device 40 controls the conveying means 20 to cause the recording medium to move past the readhead 30 more quickly, so that the number of frames given to the output 101 per unit time is the nominal value described above. To be equal to

If the user wants to select a playback speed other than the nominal speed, the user drives the selector switch 50 to input a speed factor ∝. In the following, the speed factor ∝ is defined as a factor that the nominal reproduction speed should be multiplied, where the speed factor 큰 greater than 1 corresponds to fast motion playback, and the speed factor 있는 between 0 and 1 corresponds to slow motion playback.

According to an important aspect of the present invention, the speed factor ∝ can be continuously adjusted within certain limits. In another variant, it is also possible to allow the user to select only the speed factor 으로부터 from a predetermined number of (relatively large) values.

Hereinafter, to illustrate the principle of the present invention, an example of slow motion reproduction according to the present invention will be described with reference to FIG. 2B. FIG. 2B shows a sequence consisting of a plurality of frames derived from the sequence shown in FIG. 2A by the control device 40 by repeating each B frame once. Similarly, the plurality of frames inside the new sequence are indicated by the numbers 1, 2, 3,... Shown at the top of FIG. 2B. The shape of each frame is shown at the bottom of FIG. 2B, where the original sequence number (FIG. 2A) of the frame is shown in parentheses.

Thus, in the present embodiment, a new sequence consisting of 15 frames is derived from the original sequence containing 9 frames. An important feature is that this new sequence fully follows the MPEG format, so that the controller 40 can supply this modified sequence directly to the standard display, which can directly decode and display this modified sequence. Is that there is. Since the reproduction time of the modified sequence (15 frames) is longer than that of the original sequence (9 frames), slow motion reproduction is performed. As a result, in this case, the speed factor ∝ is 9/15 or 0.6.

It is apparent that there is a larger number of "clean" speed factors 작은 less than one. In the sequence of GOPs shown for example in FIG. 2A, the B frames inside the GOP are repeated one, two, three, four or five times, thereby providing 9/10, 9/11, 9/12, It is possible to get arguments of 9/13 and 9/14. Different fractions can be obtained by considering more and more contiguous multiple GOPs. In the case of considering two consecutive GOPs, for example, two and three B frames inside the first and second GOPs, respectively, By repeating, a speed factor of 18/23 can be obtained.

It should be noted that a very small slow motion factor can be obtained by repeating the B frame one or more times. For example, the B frames of the GOP shown in FIG. 2A are each transmitted four times in succession, in which case the resulting slow motion factor is 9/27 = 0.33. However, for an even smaller number of slow motion factors, the reproduced picture may provide a feeling of sudden movement in the viewer.

In practice, the length of multiple GOPs need not necessarily be constant, so it is not certain whether a given desired speed factor can be realized in any GOP. However, this is not important, and an important feature of the present invention is that in the case of delayed reading of a plurality of frames, the nominal frame number at the output 101 can be obtained on average by repeating the plurality of B frames.

Reference will now be made to FIG. 3. The control device 40 is configured to drive the conveying means in the slow motion regeneration mode at a speed approximately equal to a multiple of the nominal speed. As a result, the average number of frames read per unit time by the readhead 30 is approximately equal to a multiple of the nominal value. As mentioned above, the control device 40 is configured to monitor the data stream read from the readhead 30 and to count the number of frames transmitted per unit time. Each time the counted value is found to be smaller than the nominal value, the controller supplies to output 101 the X times the B frame data. For this purpose, the control device 40 has a memory 41 for storing B frame data.

In principle, X is chosen to be 2. However, if the selected rate factor is so small that it is not appropriate to supply twice the B frame data to maintain the nominal frame rate, controller 40 increments X to 3 or even 4.

This configuration allows the average number of frames supplied to the output 101 per unit time to be approximately equal to the nominal value described above. Note that since the average number of frames in the modified sequence is always equal to the nominal value of the frame and the modified sequence follows the MPEG format, no modification is necessary to receive the modified signal on the display side. Should be. The display device decodes each frame received in a "normal" manner, so that it is independent of the source, i.e. whether it is a "normal" or a "repeated" frame, nominal time (1 in PAL format). / 25 seconds) on the screen.

Furthermore, it should be noted that, in principle, since there is no need to repeat the same number of B frames within each GOP, any desired slow motion factor can be obtained, clearly within reasonable limits. It is important that for any read rate (playback rate of the recording medium), the B frames are repeated at a constant moment so that the total number of transmitted plural frames, averaged over a longer time, is always approximately equal to the nominal value. Thus, the actual length of the GOP is irrelevant.

Hereinafter, to illustrate the principle of the present invention, an example of fast motion reproduction according to the present invention will be described with reference to FIG. 2C. FIG. 2C shows a sequence composed of a plurality of frames derived from the sequence shown in FIG. 2A by the control device 40 by skipping every second B frame. Similarly, the plurality of frames inside the new sequence are numbered 1, 2, 3,... The shape of each frame is shown at the bottom of FIG. 2C, where the original sequence number (FIG. 2A) of the frame is shown in parentheses.

Thus, in this embodiment, a new sequence consisting of six frames is derived from an original sequence comprising nine frames. An important feature is that this new sequence fully follows the MPEG format, so that the controller 40 can supply this modified sequence directly to the standard display, which can directly decode and display this modified sequence. Is that there is. Since the reproduction time of the modified sequence (six frames) is shorter than the original sequence (nine frames), fast motion reproduction is performed. As a result, in this case, the speed factor ∝ is 9/6 or 1.5.

It is apparent that there is a larger number of "clean" speed factors 보다 greater than one. In the sequence of GOPs shown for example in FIG. 2A, by skipping one, two, three, four, five or six B frames within the GOP, 9/8, 9/7, 9 It is possible to obtain factors of / 6, 9/5, 9/4 and 9/3. A different fraction can be obtained by considering more and more contiguous multiple GOPs, in which case two and three B frames inside the first and second GOPs are skipped, for example when considering two consecutive GOPs. By doing so, a speed factor of 18/13 can be obtained.

It is obvious that the maximum fast motion factor that can be obtained depends on the number of frames in the GOP. Also, a larger fast motion factor can be obtained by skipping multiple P frames, but this is only allowed if all B frames associated with such P frames are likewise skipped. If all P frames are skipped, only one frame of the GOP is indicated in this embodiment, with the resulting fast motion factor being 9/1.

Reference will now be made to FIG. 3. The control device 40 is configured to drive the conveying means in the fast motion regeneration mode at a speed approximately equal to a multiple of the nominal speed. As a result, the average number of frames read per unit time by the readhead 30 is approximately equal to a multiple of the nominal value. As mentioned above, the control device 40 is configured to monitor the data stream read from the readhead 30 and to count the number of frames transmitted per unit time. Each time the counted value is found to be greater than the nominal value, the controller 40 blocks the B frame data (and, if necessary, the data of the P frame as well), so that this data cannot reach the output 101. do.

This configuration allows the average number of frames supplied to the output 101 per unit time to be approximately equal to the nominal value described above. In this case, since the average number of frames in the modified sequence is always equal to the nominal value of the frame, and furthermore, since the modified sequence follows the MPEG format, no modification is necessary to receive the modified signal on the display side. It should be noted that The display device decodes each frame received in the "normal" manner and reproduces it on the screen for a nominal time (1/25 second in PAL format).

Moreover, it should be noted that, in principle, since there is no need to skip the same number of B frames within each GOP, it is possible to obtain any desired fast motion factor, clearly within reasonable limits. For any read rate (playback rate of the record carrier), it is important that the B frames are skipped at certain moments so that the total number of a plurality of transmitted frames, averaged over a longer time, is always approximately equal to the nominal value. Thus, the actual length of the GOP is irrelevant.

An important advantage of the method presented according to the invention is that the reproduced picture does not exhibit any disturbance, and that the motion component of the picture actually exhibits faster movement. As mentioned above, the maximum fast motion factor obtainable by the method according to the invention depends on the number of B frames (P frames if possible) within the GOP. Clearly, by skipping I frames, a larger fast motion factor can be obtained. However, at such high playback speeds (9 or more fast motion factors), this possibility provided by the present invention has little practical advantage since the viewer cannot follow a faster moving picture accordingly.

The present invention provides an alternative to generate a more stable picture for the viewer. According to this modification, first, the predetermined first number of consecutive frames is transmitted to the display device, and then the predetermined second number of consecutive frames is skipped. In this case, it is preferable that the first number and the second number correspond to the total number of the plurality of GOPs. In the example where nine consecutive GOPs are skipped after one GOP is transmitted, the resulting fast motion factor is 10/1.

The advantage of this fast motion playback method presented by the present invention is that the viewer can always see the moving picture scenes normally moving for a short time, and the continuous picture scenes are separated from each other by a jump in time. Since the length of this time jump is hardly or impossible to estimate, in this respect watching the playback using the fast motion factor of 5 is similar to watching the playback using the fast motion factor of 50.

The scope of protection of the present invention as defined in the claims is not limited to the embodiments shown in the accompanying drawings and described herein, but is not limited to the foregoing description of the apparatus and method according to the invention within the scope of the present invention. It will be apparent to those skilled in the art that modifications or variations of the embodiments are possible.

Claims (5)

A method of reading a sequence consisting of a plurality of video frames coded according to the MPEG format and comprising a plurality of I frames, P frames, and B frames, wherein the recording medium provided with the recording material is reproduced at a speed different from the nominal reproduction speed. The plurality of video frames of the sequence are supplied at a speed different from the nominal speed (the number of frames per unit time), and one or more B frames are repeated or skipped among the plurality of frames thus read, thereby providing a plurality of frames per unit time. And transmitting the modified sequence to the display device in which the average number of frames is equal to a predetermined nominal value. The method of claim 1, And the recording medium is reproduced at a speed higher than the nominal reproduction speed, and in addition to the plurality of B frames, the plurality of P frames is skipped. The method of claim 1, The recording medium is reproduced at a speed higher than the nominal reproduction speed, and after each predetermined first number of consecutive frames is transmitted to the display device, each time the predetermined second number of consecutive frames is skipped, the second number Has a value greater than the first number. The method of claim 3, wherein And said predetermined first number of consecutive frames corresponds to the total number of a plurality of GOPs, and said predetermined second number of consecutive frames corresponds to the total number of a plurality of GOPs. Carrying out the method according to any one of claims 1 to 4, A read head 30 for reading information recorded on the recording medium 10, Conveying means 20 for moving the recording medium 10 and the read head 30 at an adjustable speed relative to each other; An output 101 for supplying a sequence consisting of a plurality of coded frames, A control device 40 connected to the readhead 30 for receiving information to be read and connected to the output for selectively supplying the read information to the output 101; A memory 41 which temporarily stores a plurality of read B frames in association with the control device 40; In the apparatus 100, which is driven by a user and has a speed selection switch 50 connected to the control apparatus 40, The control device 40 is connected to the conveying means 20 and, in response to the signal received from the speed selector switch 50, causes the conveying speed of the recording medium 10 to be a multiple of a predetermined nominal speed. Configured to control the transfer means 20, where k is the speed factor selected using the speed select switch 50, The control device 40 is configured to monitor the data stream read from the read head 30 and compare the number of frames per unit time in the data stream with a predetermined nominal average value, When the number of frames counted per unit time is smaller than the nominal average value, the control device 40 stores the data of the B frame in the memory 41 and transmits the data to the output 101. Read from memory 41 and send the data back to output 101, The control device (40) is characterized in that the data of the B frame does not reach the output (101) when the number of frames counted per unit time is greater than the nominal average value.