JP3127975B2 - Storage / playback control method and apparatus for fast forward and reverse fast forward playback of digital video program - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the storage of a randomly accessible digital video program in a frame such as a magnetic disk at a center of a request type video program providing system and the like, and stores the digital video program in frames. The present invention relates to a method of storing and reading out a moving image program, in which a reproduced moving image program is played back normally, fast-forwarded, and reverse fast-forwarded. The invention also relates to an apparatus for performing the method.

[0002]

2. Description of the Related Art In a request-type moving picture program providing system, a center is connected to a large number of terminals, and a moving picture program requested by the terminal is read from a storage device of the center.
Sent to the terminal. As a storage medium for moving image data, a magnetic disk is mainly used. In order to increase the storage capacity and improve the data transfer speed, a plurality of magnetic disks are arrayed and simultaneous access control to each disk is performed. When reading video data from a magnetic disk,
In addition to the moving image data reading time, an access overhead time such as movement of the head and waiting for the rotation of the disk is required. In order to reduce the access overhead time and efficiently read out the moving image data, several consecutive moving image frames are stored as a group adjacently on the same track (sometimes over adjacent tracks), and the moving image data is stored. Reading is performed in group units.

In order to perform fast-forward playback of a moving picture program (the same applies to reverse fast-forwarding, the description is omitted below), moving-picture data stored in the order of frame numbers for normal playback are separated from the moving picture data required for fast-forward playback. The frame stored at the location must be read. Conventional fast-forward methods include:
There are (1) a method of repeatedly reading one frame and moving the read head of the storage device, and (2) a method of extracting only moving image frames necessary for fast-forward playback in advance and creating fast-forward-only data.

Now, a motion picture program is displayed in full motion (3.
0 [frame / s]), and the transfer speed of the moving image data is set to 6 [Mbps] (ISO standard MPEG2
(Equivalent bit rate ), the size of one moving image frame is 0.2 [Mbits]. Assuming that the access overhead time of the magnetic disk is D [ms] and the data read speed is V [Mbps], when the magnetic disk reads moving image data of continuous 30 frames (hereinafter referred to as case 1), the read is performed. Time T ¹ [ms]
Is T ¹ [ms] = D [ms] + 30 × 0.2 [Mbits] × 1000 / V [Mbps] (1) Here, assuming that D [ms] = 150 and V = 25 [Mbps], T ¹ [ms] = 150 [ms] + 30 × 0.2 [Mbits] × 1000/25 [Mbps] =
390 [ms] (2). Therefore, moving image data for one second can be read out and reproduced at 390 [ms].

[0005] When the magnetic disk is repeated and read video data of one frame, the movement of the read head (hereinafter, the case 2 hereinafter), 1 frame read time T ₂ [ms] is, T ₂ [ ms] = D [ms] + 0.2 [Mbits] x 1000 / V [Mbps] (3) As in the above case, D [ms] = 150, V = 25 [Mbps]
Then, T ₂ [ms] = 150 [ms] + 0.2 [Mbits] × 1000/25 [Mbps] = 158 [ms] (4) Therefore, in case 2, only 6 frames are read per second. Can not. As described above, when a plurality of frames are read as one group as in Case 1, the moving image data can be efficiently read. However, as in Case 2, the reading of one frame of moving image data and the movement of the read head are performed. Is repeated, the number of moving image data read per second decreases. The above method 1 corresponds to case 2.

FIG. 7 is a block diagram showing an example of the configuration of a conventional fast-forward and reverse-fast-forward playback control system. This fast-forward playback control system includes a terminal 71, a network 7
2. Network interface unit 73, switch unit 7
5, a read management unit 74, a storage device control unit 76, and a storage device 77.

In FIG. 7, a terminal 71 includes a communication control unit,
It is composed of a decoder, AV equipment, and input equipment, transmits a user request specified by the input equipment to the center device, and decodes (decompresses) and displays moving image data received from the center device. The network 72 is, for example, an ISDN
(Integrated Services Digital Network) network. The network interface unit 73 receives a request for reading moving image data from the terminal 71 via the network and transmits moving image data. The read management unit 74
Each terminal 7 includes a communication control unit and a microprocessor.
1 and a read processing instruction to the switch unit 75. The switch unit 75 is, for example, a VME (Versa Module Europe) bus, and configures a transmission path between the storage device control unit 76 and the network interface unit 73. The storage device control unit 76
For example, a magnetic disk control board, for example, SCSI (Sma
In the case of the ll Computer System Interface) interface, a maximum of seven storage devices can be daisy-chained to one storage device control unit to expand the storage capacity. The storage device 77 is, for example, a magnetic disk.
When each user requests a desired moving picture program from each terminal 71, the request reaches the readout management unit 74 in the center device via the network 72. The read management unit 74 searches for a storage device (usually provided in the read management unit) in which the directory of the requested moving image program is stored, and instructs the storage device control unit 76 to read.

[0008]

In the fast-forward and reverse-fast-forward playback control system shown in FIG. 7, the fast-forward playback method (1) for a moving picture program described above uses data seek and rotation waiting every time one frame is read during fast-forward playback. Head access overhead occurs. As a result, the amount of data read within one slot time decreases, and as a result,
There is a problem that the reproduced fast-forward image has awkward frames. In the fast-forward playback method (2) for a moving picture program described above, moving-picture data for fast-forward playback must be prepared in advance separately from movie data for normal playback. There is an additional problem, and when the playback mode is changed between the normal playback and the fast-forward playback, it is necessary to control such that the moving picture program transitions between the two pieces of moving picture data without interruption. Furthermore, there is a problem that the double-speed value of fast-forward is uniquely fixed by creating moving image data for fast-forward playback in advance.

An object of the present invention is to eliminate the need to separately prepare normal reproduction data and fast-forward reproduction data, and to increase the number of reproduction frames per unit time during fast-forward reproduction to smooth the motion of a moving image. Therefore, it is possible to provide a digital moving image storage / playback control method and a device for implementing the method, which can read moving image data efficiently and flexibly change the double speed value of fast forward. is there.

[0010]

According to the digital moving image storage / reproduction control method of the present invention, a digital moving image is stored in a plurality of storage devices in frame units, and reproduced in response to a reproduction request instructed from a terminal. This is a moving image storage / playback control method. In this method, for a number m of a plurality of storage devices, moving image frames having the same frame number modulo m are adjacent to the same storage device in the order of frame numbers. Accumulate.

At the time of normal reproduction, a predetermined number N of moving image frames stored in each storage device in the order of frame numbers are set as a reading unit, and moving image frames included in the reading unit are sequentially read and buffered. The processing is executed in parallel with respect to the plurality of storage devices, mN moving image frames read from the plurality of storage devices and buffered are reproduced as a reproduction unit, and the reading and buffering of the moving image frames and buffering are performed. The playback of the ringed moving image frame is repeated in the order of the reading unit of the moving image frame stored in each storage device, and at the time of fast forward playback, the playback frame number at the time of the fast forward playback request is set as the first term. In order to form a sequence of the second term or less of the arithmetic progression having a double speed coefficient n set to a relatively prime value as a tolerance. The fast-forward frame number is specified, the identification number of the storage device storing the moving image frame of the specified fast-forward frame number is calculated, and the moving image frame of the specified fast-forward frame number is calculated from the storage device of the calculated identification number. When the processing of reading and buffering and reading and buffering the moving image frame has completed all the storage devices, the moving image frame buffered is reproduced as a reproduction unit.

In order to facilitate the setting of the double speed coefficient for fast-forwarding, the number m of storage devices is preferably a prime number. The fast-forward playback includes reverse fast-forward playback, and when a reverse fast-forward playback request is made, the double speed coefficient can be set to a negative number.

It is desirable that the remainder obtained by dividing the frame number of the moving image frame stored in each of the plurality of storage devices by m be determined as the storage device number of the storage device. After the storage device numbers are determined in this way, moving image frames can be stored in the order of the storage device numbers and the frame numbers. It is desirable that the number N of frames in the reading unit is determined such that mN is equal to the number of frames per second in full motion display.

The digital moving image storage / playback control device of the present invention is a digital moving image storage / playback control device that stores a digital moving image in a storage device on a frame basis and plays it back in response to a playback request. The device includes a plurality of storage devices arranged in parallel, and a plurality of storage devices each including a buffer memory for controlling reading / writing of the storage device and buffering moving image data read from the storage device. A storage control device having a device control unit; an interface unit for interfacing between the terminal and the storage control device via a network; a data line for transmitting moving image data read in parallel from a plurality of storage devices; A switch unit for switching the connection between the ace units, receiving a reproduction request signal and a reproduction mode change request signal transmitted from the terminal,
A read management unit that instructs the storage control device to read the moving image data and buffer the read moving image data, and instructs to play back the buffered moving image data; Each storage device of
For the number m of the plurality of storage devices, a moving image frame having a congruent frame number modulo m is stored in an adjacent position in the order of the frame number, and the read management unit, when a normal reproduction request is received from the terminal, The playback frame number at that time is 1
Is incremented by one, and the identification number of the storage device storing the moving image frame of the incremented frame number is calculated.
Instructs the storage control device to sequentially read and buffer the moving image frames in the order of the frame numbers, and then instructs the storage control device, then increments the frame number by one, and stores the moving image frame of the incremented frame number. Calculate the identification number of the storage device performing the operation, instruct the accumulation control device to read / buffer the moving image frame, and repeat the read / buffering instruction until all storage devices have made one round. A normal playback using mN moving image frames as a playback unit, which is read in parallel from each of the storage devices in response to a read / buffering instruction executed in a loop over all the storage devices, and a storage control device. Instructs the switch section, and the read management section further performs a fast-forward playback request from the terminal when the playback is requested. The frame number is incremented by m by a double speed coefficient n set to a relatively prime value, and the identification number of the storage device storing the moving image frame of the incremented frame number is calculated. The storage control device instructs the storage control device to read and buffer the moving image frame from the storage device of the number, and then increments the frame number by n, and stores the moving image frame of the incremented frame number in the storage device. The identification number is calculated, the reading and buffering of the moving image frame is instructed to the storage control device, and the reading and buffering instructions are repeated until all the storage devices are cycled. Reads from each storage device in parallel in response to the read and buffering instructions performed in one round In addition, the storage control device and the switch unit are instructed to perform fast-forward playback using m moving image frames as a playback unit, and the above readout, buffering instruction, and fast-forward playback instruction are repeated. When a normal playback instruction or a fast-forward playback instruction is received, the moving image data buffered in each storage device control unit is read onto a data line between the storage device control unit and the switch unit, and the switch unit is read out. In order to transfer the moving image frame to the terminal, a signal path between the storage device control unit and the interface unit is set according to the control of the read management unit.

[0015]

First, the principle of the present invention will be described. First, in order to increase the number of playback frames per unit time at the time of fast-forward playback, which is one of the above-mentioned objects, the present invention intends to read moving image data in parallel from m (plural) storage devices. Have been. For this purpose, moving image frames having the same frame number modulo m are stored in the same storage device. Now, the identification number of the storage device is D ₀ ,
D _1, D _2, When _{···· D m-1, D 1} : 1,1 + m, 1 + 2m, ····, D 2: 2,2 + m, 2 + 2m, ····, D 3: 3, 3 + m, 3 + 2m, ····, · (5) · · D m-1: m-1,2m-1,3m-1, ····, D 0: m, 2m, 3m, ···· . Here, a and b are congruent modulo m if ab
Is divisible by m. D ₀ , D ₁ , D _2.
.. D _m-1 is called a coset modulo m in algebra, and any integer belongs to any of these classes. Were it to, the frame number of video frames categorized as described above, a video frame corresponding to each frame number, storage unit number _{D 0, D 1, D 2} , ··· D m-1
Of the moving image frames stored in the same storage device, the relationship of the frame numbers between the moving image frames stored in different storage devices, the storage device number D Since the relationship with the frame number of the moving image frame stored in the storage device is extremely simple and clearly defined, as shown in the embodiment described below, a plurality of storage The device can be accessed.

The present invention is configured on the assumption that a moving image frame is allocated to each storage device as described above. At the time of normal reproduction, N moving image frames are read in parallel from all the storage devices as a read unit for each storage device. As can be seen from equation (5) above, the storage device D ₀
The frame numbers of the moving image frames stored in .about.D _m-1 are an arithmetic progression having a tolerance m whose first term differs by one. Therefore, the moving image frames that are read in parallel from all the storage devices only by an instruction to read N frames uniformly in each storage device are mN moving image frames having consecutive frame numbers. Therefore, by reproducing the mN moving image frames as a reproduction unit, correct reproduction can be realized without generating a missing frame or intervening an extra frame. The reading and buffering of the mN moving image frames are repeated every time allotted to the terminal to reproduce the mN frames.

The reading time is reduced by the normal reproduction according to the method of the present invention for the following reasons. a. Reading is performed in parallel; b. Reading from each storage device is performed continuously for N frames (Equation 3).
reference). Among them, b. The effect due to the above reason is greater as the length N of the readout unit is longer. However, the length of N is
It is determined depending on the time allocated to the terminal to play mN moving image frames. Usually, m
N is determined to be equal to the number of frames per second in full motion display. However, no matter how long N is set, the advantage of the present invention that the above-mentioned correct reproduction can be realized is not lost.

The present invention operates as follows in fast-forward playback. In the case of fast-forward playback, it is important how to allocate moving image frames to a plurality of storage devices and which of the moving image frames stored in each storage device is selected as a reproduction frame. The reason is that, in the case of fast-forward playback, among the moving image frames stored in each storage device, a skipping moving image frame is selected as a reproduction frame, so that the method of allocating the moving image frame to a plurality of storage devices and Depending on the manner in which a playback frame is selected from moving image frames stored in each storage device, a playback frame may be selected from only a specific storage device. In such a case, the reading of the moving image frame from the specific storage device becomes a rate-determining process, and even if a plurality of storage devices are used, the effect of reading the moving image frame from those storage devices in parallel cannot be expected. As a simple example, a case in which moving image frames with even frame numbers are stored in one of the two storage devices and moving image frames with odd frame numbers are stored in the other storage device to perform double-speed fast-forward reproduction. In this case, since only the moving picture frame having the even or odd frame number is selected, only one storage device is read, and the effect of using two storage devices is lost.

In the present invention, as shown in equation (5), after allocating moving image frames to the storage devices, a double speed coefficient n set to a relatively prime value with respect to the number of storage devices m To fast forward playback. In this way, by setting n and m to be relatively prime, after the moving image frame of the frame number f stored in a certain storage device D is selected as the fast-forward playback frame, fast-forwarding is performed m times. The playback frame is selected, and the frame number f + m
n (mn is the least common multiple of the number m of storage devices and the speed factor n)
Until is selected, the moving image frames stored in the storage device D are not selected. This means that m
This holds for each of the storage devices. Therefore, m playback frames that are sequentially selected should belong to different storage devices. In other words, m
If one of the playback frames is selected, the storage device of m pieces will be read once by going through the storage device once.
It is avoided that data is read out more times than other storage devices. Thus, the present invention intends to achieve read efficiency in proportion to the number m of storage devices by ensuring that the same number of video frames are read from each storage device at the same time.

The operation described above is clearly shown in FIG. 5, for example. In the case of FIG. 5, m = 5 and n = 3. It goes without saying that n and m are in a prime relationship with each other. After being frame number 1 is selected from the magnetic disk 7 ₀ until the frame number 16 (= 1 + 15) is selected, triple-speed fast-forward reproduction frame 4,7,10,13 is, after searching different magnetic disk Is read. The moving image frames read out before making a complete round of all the magnetic disks are correct 3 × fast-forward moving image frames that include necessary moving image frames without omission and do not include unnecessary moving image frames. Therefore, by transmitting a fast-forward moving image frame obtained by looping over all magnetic disks to the terminal as a playback unit, correct fast-forward playback can be achieved. As described above, in order to achieve correct fast-forward reproduction according to the present invention, it is necessary that the double speed coefficient n and the number m of storage devices are mutually prime. This means
It is easily realized by taking m as a prime number.

As described above, the relationship between the frame numbers of the moving image frames stored in the respective storage devices, the relationship between the frame numbers of the moving image frames stored in the different storage devices, the storage device number and the storage in the storage device. Since the correspondence with the frame number of the moving image frame is simple and clear, the designation of the fast-forward frame number by a simple program, the calculation of the storage device number corresponding to the fast-forward frame number, and the calculation of the storage device number From the corresponding storage device, sequentially read out and buffering instructions of the moving image frame of the specified fast-forward frame number,
In response to the instruction, the specified moving image frame is read from the specified storage device. At this time, the time required for the designation of the fast-forward frame number, the calculation of the storage device, the readout, and the instruction of the buffering is negligibly small as compared with the readout time of the moving image frame. Read in parallel. Reverse fast-forward playback can be performed in exactly the same procedure as fast-forward playback except that the double speed coefficient n is set to a negative value. By determining the storage device number to be equal to the remainder obtained by dividing the frame number of the moving image frame stored in the storage device by m, the storage device number and the frame number of the moving image frame stored in the storage device are determined. The correspondence between them can be more clearly defined. As a result, by simply storing moving image frames in a plurality of storage devices in the order of storage device numbers and frame numbers, moving image frames having the same frame number can be stored in the same storage device by using m as a modulus.

The digital moving picture storage / playback control apparatus of the present invention is an apparatus for implementing the digital moving picture storage / playback control method of the present invention. Therefore, the operation is the same as that of the digital moving image accumulation / reproduction control method of the present invention.

According to the present invention, moving image data is individually read from the storage devices configured in parallel.
Even in the case of reading out moving image data stored physically apart from each other, it becomes possible to efficiently read out moving image data by the number of parallel storage devices. As a result, the number of display frames per unit time can be increased at the time of fast-forward playback, and as a result, smooth fast-forward playback becomes possible. It is not necessary to prepare data for normal reproduction and data for fast forward separately, and there is no need to transition between these two types of moving image data during fast forward reproduction. By appropriately adjusting the frame number for reading a moving image frame from each storage device, it is possible to flexibly change the double-speed value of fast-forward.

[0024]

FIG. 1 is a block diagram of a storage / reproduction control system according to an embodiment of the present invention. The storage / reproduction control system according to the present embodiment includes a terminal 1, a network 2, and a storage / reproduction device (hereinafter, referred to as a center unit). The center unit includes a network interface unit 3, a read management unit 4, a switch unit 5, A storage device controller 6 and a storage device 7 are provided. In the following description, the storage device controller 6 and the storage device 7 are collectively referred to as a storage and playback unit. As the storage device 7,
Although a magnetic disk, a magneto-optical disk, a semiconductor disk, and the like can be used, this embodiment is an example using a magnetic disk. The storage device controller 6 is, for example, a magnetic disk control board, and a plurality of storage devices (5 in this embodiment) are provided in the center.
Clause) is provided. FIG. 1 clearly shows three of the five items. Each storage device controller 6 controls the corresponding storage device 7 to write and read image data. The storage device connected to one storage device control unit includes:
For example, SCSI (Small Computer System Interface)
In the case of an interface, up to seven magnetic disks 7 can be connected in a daisy chain to extend the storage capacity. The switch unit 5 is, for example, a VME (Versa Mo
dule Europe) bus, and constitutes a transmission path between each storage device control unit 6 and the network interface unit 3 under the control of the read management unit 4. Read management unit 4
Is composed of a communication control unit and a microprocessor, receives a reproduction request from each terminal 1, gives a read processing instruction to the storage device control unit 6 according to the reproduction request, and transmits the read moving image data to the terminal 1. , The connection of the switch unit 5 is controlled. The network interface unit 3 interfaces between the terminal side and the storage / reproduction unit via the network 2, and receives a reproduction request for moving image data from the terminal 1 and transmits moving image data. The network 2 is, for example, an ISDN (Integrated S
ervices Digital Network) network. The terminal 1 includes a communication control unit, a decoder, an AV device, and an input device. The terminal 1 transmits a reproduction request of a user specified by the input device to the center, and decodes (decompresses) moving image data received from the center. To display.

FIG. 2 is a diagram schematically showing a method of accumulating moving image data in frame units according to the present embodiment. FIG.
, The number m of parallel magnetic disks is 5. Full-motion display of video programs (30 [frame / s])
If the transfer speed of the moving image data is 6 [Mbps] ( bit rate equivalent to MPEG2 in the ISO standard), one moving image frame size is 0.2 [Mbits]. Assuming that the access overhead time of the magnetic disk is D [ms] and the data read speed is V [Mbps], when the storage device controller 6 reads moving image data for six consecutive frames (hereinafter referred to as case 3), , Read time T ³
[Ms] is T ³ [ms] = D [ms] + 6 × 0.2 [Mbits] × 1000 / V [Mbps] (6) Now, D = 150 [ms], V = 25
Assuming [Mbps], the read time of Case 3 is given by the following equation.

T ₃ [ms] = 150 [ms] + 6 × 0.2 [Mbits] × 1000/25 [Mbps] = 198 [ms] (7) During normal reproduction, five magnetic disks configured in parallel Reads out six frames of moving image data during 198 [ms], and reads out moving image data of a total of 30 frames (for one group). This is a moving image data amount for displaying a moving image program for one second. Also, at the time of fast-forward playback, moving image data stored at a distant position on the magnetic disk is read by repeating the movement of the magnetic disk read head and the reading of one frame. corresponding to the read time T ₂ of the case 2. Therefore, the read time T ₂ are made in equation (4) approximately 160 [ms] as represented by the number of frames to be read per second is about 6. As a result, moving image data of 30 frames (for one group) per second is read from the five magnetic disks configured in parallel. This is a data amount for displaying a moving image program for one second.

In the request type moving picture program providing system, multiple access control to the magnetic disk is performed so that a plurality of users can simultaneously access the same magnetic disk. The technique utilizes the fact that the data reading speed of a magnetic disk is different from the decoding (expansion) speed of moving image data, and time-divides the access time to the magnetic disk and allocates it to a plurality of users. The access to the magnetic disk is performed periodically, and the access time to the magnetic disk in each cycle is a certain unit time (hereinafter, referred to as a slot).
And assign one user to each slot. Each slot is equal to the sum of the access overhead time and the reading time of one group of moving image data (or
Greater). When the time of one slot is short,
If you perform the reading method like Case 2 for fast forward,
The number of frames to be read further decreases, resulting in an awkward image. In the present embodiment, the number of frames to be read increases in proportion to the number of parallel magnetic disks, so that there is an effect that a moving image can be approximated.

Next, the operation of the read management section 4 of this embodiment will be described with reference to the flowchart shown in FIG.

In FIG. 3, a symbol f represents a playback frame number at the start of reading a moving picture program or at the time of changing the playback mode.
The symbol m is the number of parallel magnetic disks, and the symbol n is a double speed coefficient (± 2, ± 3, ± 4,...) For fast forward or reverse fast forward playback.
Take the value of In some cases, a value other than an integer multiple is also possible. , Where a positive value indicates a fast forward value and a negative value indicates a double speed value of a reverse fast forward. i and j are loop counters in the flowchart.

First, at the start of reading a moving image program, the reproduction frame number f is initialized to 0, the loop counters i and j are initialized to 0, and the number of magnetic disk devices m = 5 is set (S).
1). Next, check the playback mode.
Go to (S2). In the case of normal reproduction, the moving image frame number being reproduced at that time (at the start of reading the moving image program or when the reproduction mode is changed) is substituted into f (S3). The reproduction frame number f is incremented by 1 (S4). The number D of the magnetic disk storing the moving image frame of the frame number f is calculated by the following equation.

D = (f + m−1) mod m (8) where A mod B represents a remainder obtained by dividing A by B. Frame f from the magnetic disk 7 _D, f +
m, f + 2m, f + 3m, f + 4m, read and read buffering of video data f + 5 m instructs the storage device controller 6 _D corresponding to the magnetic disk (S5). The loop counter i is incremented by 1 (S6). The determination for instructing the reading of the moving image data by the number of magnetic disks is performed. If there is a magnetic disk for which reading has not been instructed yet (YES), S
If the reading instruction has been issued to all the magnetic disks to NO. 4 (NO), the flow proceeds to step S8 (S7). In step S8, the loop counter i is initialized to 0 (S
8). Next, normal reproduction of the buffered moving image data is instructed (S9). The specific content of this playback instruction is
Read control of moving image data buffered in each of the storage device control units 6 ₀ , 6 _1, 6 ₄ , and communication control for switching a switch unit to transmit the read moving image data to the terminal that has requested reading. It is. Next, it is determined whether or not to continue the normal reproduction in response to the read request. In the case of continuation (YES), the flow proceeds to S11, and in the case of an instruction to change the reproduction mode (NO), the flow proceeds to S2 (S10). After 5m is added to f in step S11, the process proceeds to step S4.

In the case of fast-forward or reverse-fast-forward playback in S2, the number of the moving picture frame being played at that time is substituted for f, and the speed factor is set to n (S12). The reproduction frame number f is incremented by n (S13). The magnetic disk number D is calculated from the magnetic disk D = (f + m-1) mod m, and the reading of the frame f and the buffering of the read moving image data are instructed (S14).
The loop counter j is incremented by 1 (S1
5). A determination is made to instruct reading of moving image data by the number of magnetic disks. If there is a magnetic disk for which reading has not been instructed yet (YES), S13
If the reading instruction has been issued to all the magnetic disks (NO), the process proceeds to S17 (S16). The loop counter j is initialized to 0 (S17). The fast forward or reverse fast forward playback of the buffered moving image data is instructed (S18). In this case as well, the specific contents of the reproduction instruction are the same as those in step S9, but step S1
The reproduction of step S9 differs from the reproduction of step S9 in that reproduction is performed in the positive and reverse directions according to the sign of the double speed coefficient n. Next, it is determined whether or not to continue the fast forward or reverse fast forward playback. In the case of continuation (YES), the flow proceeds to S13, and in the case of an instruction to change the reproduction mode (NO), the flow proceeds to S2 (S19). Hereinafter, the same operation as described above is repeated.

Referring to the storage state of the moving image data in FIG.
The normal reproduction of the flowchart of FIG. 3 will be described. Now, when a request to change the playback mode from fast-forward playback to normal playback is sent from the terminal to the read management unit, the read management unit 4
Receives the frame number f at that time from the terminal at the same time (S3). Now, assuming that f = 32, normal reproduction is started from f = 33 (S4). The read management unit 4 sets f =
33 calculates the magnetic disk number D = 2 accumulated, the frame number from the magnetic disk 7 ₂ 33,38,4
3, 48, 53, and 58, and instructs buffering of the read moving image data (S
5), the loop counter i is incremented and i = 1
(S6). If it is determined that i = 1 is smaller than the number m of magnetic disks (S7), the process jumps to step S4 and sets f to 34. Next, a magnetic disk number D = 3 in which the frame of the frame number 34 is stored is calculated,
Frame number 34,39,44 from the magnetic disc 7 _3,
It instructs the reading of the frames 49, 54 and 59 and the buffering of the read moving image data (S5). The loop counter is incremented to 2 (S6), and when it is determined that the loop counter value 2 is smaller than m = 5 (S6).
7) Jump to step S4. Thereafter, the same operation is repeated, and the frame numbers (35, 40, 45, 50, 55, 6) are respectively sent to the storage device controllers 6 ₄ , 6 ₀ , 6 ₁ .
0), (36, 41, 46, 51, 56, 61), (3
7, 42, 47, 51, 57, and 62). In this way, when all the magnetic disks 7 ₀ , 7 ₁ , 7 ₂ , 7 ₃ , 7 ₄ have been cycled and the reading of 6 frames and the instruction of buffering have been completed, a total of 30 frames (1 The reading instruction and the buffering instruction of the moving image frame for one second in the full motion display are completed. Reading management unit 4 resets the loop counter i in 0 (S8), and instructs the normal reproduction for each storage controller 6 ₀ ~6 ₄ (S9), when there is no request for changing the playback mode from the terminal Sets f = 62 (S11), and then goes to S4 to issue a read instruction and a buffering instruction for the next one group by the same procedure. When the terminal requests the change of the reproduction mode, the procedure performed by the read management unit 4 jumps to step S2.

FIGS. 4, 5, and 6 show examples of frame numbers of moving image data read from each magnetic disk 7 when fast-forward playback is performed according to the flowchart of FIG. A case where fast forward reproduction is started from frame number 1 is shown. Data read by each disk is indicated by hatching.

FIG. 4 shows the case of double speed reproduction,
After request, magnetic disk 7₀From frame numbers 1, 11,
21 ..., magnetic disk 7₁From frame number 7, 17,
27 ..., magnetic disk 7_TwoFrom frame numbers 3, 13,
23, magnetic disk 7_ThreeFrom frame numbers 9, 19,
29, magnetic disk 7_FourFrame numbers 5, 15,
25 are read out. FIG.
Is the case of 3x speed playback, after a fast forward request,
Screen 7₀, Frame numbers 1, 16, 31, ..., magnetic
Screen 7₁, Frame numbers 7, 22, 37 ...
Screen 7_Two, Frame numbers 13, 28, 43 ...
Disk 7_Three, Frame numbers 4, 19, 34 ...
Disk 7_FourVideo with frame numbers 10, 25, 40 ...
The frame is read. FIG. 6 shows the case of 4 × speed reproduction.
Yes, after fast-forward request, magnetic disk 7 ₀From frame number
No. 1, 21, 41..., Magnetic disk 7₁From frame number
No. 17, 37, 57, magnetic disk 7_TwoFrom the frame
Numbers 13, 33, 53 ..., magnetic disk 7_ThreeFrom
, The magnetic disk 7_FourFrom
The video frames with the program numbers 5, 25, 45,.
You.

Next, the procedure executed by the read management section 4 to perform the 3 × fast forward reproduction shown in FIG. 5 will be described with reference to the flowchart of FIG. As described above, the read control unit 4 receives the playback frame number f and the double speed coefficient n at that time together with the fast-forward request signal from the terminal. FIG. 5 shows an example in which fast-forward playback is started from frame number 1, so that f + n in step S13 in FIG. Reading management unit 4, in step S14, and calculates the magnetic disk number D = 0 corresponding to the frame number 1, the storage device control unit an instruction to buffer reads the video frame of the frame number 1 from the magnetic disk 7 ₀ 6 _A transmission is made for ₀ , and the loop counter is incremented to 1 (S15). Since the loop counter value 1 is smaller than the number m of magnetic disks (S16), the process jumps to step S13, and sets the frame number to f + n = 4 (S13). It calculates the magnetic disk number D = 3 corresponding to the frame number 4, the storage device control unit 6 ₃ corresponding to the number 3, and instructs the reading and buffering the video data (S14). Next, the loop counter is incremented. This procedure is repeated, and the storage device controllers 6 ₁ , 6 ₄ , 6 ₂ are sequentially assigned frame numbers 7, 7, respectively.
The reading and buffering of the moving image frames 10 and 13 are instructed. When the reading and buffering instructions have cycled through all the storage device control units, the loop counter is reset to 0 (S17), and the storage device control units _60- .
6 ₄ and fast forward to the switch unit 5, instructs the inverse fast forward reproduction (S18). The read management unit 4 determines whether or not there is a request for changing the reproduction mode from the terminal (S19), and if there is no request, repeats the above procedure. When there is a request to change the reproduction mode, the process jumps to step S2.
Although the reading and buffering instructions to the storage device control unit described above are sequentially performed, the time required for the instruction to make a round through all the storage device control units is shorter than the reading time of one frame (about 100 ms). hand,
Short enough to ignore. Therefore, the reading of the moving image frame from each magnetic disk by each storage device control unit is performed almost simultaneously. Therefore, the number of moving image frames read out per unit time increases in proportion to the number of magnetic disks. As described above, by appropriately selecting the number of moving image data read from each magnetic disk, it is possible to flexibly change the double-speed value of fast-forward reproduction. However, if the double speed value of fast forward or reverse fast forward is a multiple or a divisor of the number m of magnetic disks, reading is performed only from a specific magnetic disk. I can't get it. Further, depending on the performance of the magnetic disk (access overhead time (D) of the magnetic disk, data read speed (V)) and the configuration of the apparatus (the number of parallel magnetic disks (m)), the double speed value of the fast forward or reverse fast forward is an integer. Sometimes it does not. In this case, it is possible to adopt a method such as changing the method of designating the speed coefficient according to the performance or the device configuration, or adopting a reading method that provides a value closest to the designated speed coefficient.

[0037]

As described above, according to the present invention, for a plurality of storage devices m, moving image frames having the same frame number modulo m are stored adjacently in the same storage device in the order of frame numbers. By doing so, the following effects can be obtained. 1. During normal playback, the same number N of video frames are called in parallel from each of the m storage devices, so that mN video frames having consecutive frame numbers without unnecessary video frame intervention and necessary video frame omissions. Can be read, and as a result, the efficiency of reading moving image data can be improved. At the time of fast-forward playback, an arithmetic progression in which the speed factor n is set to be relatively prime to the number m of storage devices, the playback frame number when a fast-forward playback request is made is the first term, and the speed factor n is a tolerance , The fast-forward frame number is sequentially designated so as to form a sequence of the second term or less, the identification number of the storage device storing the moving image frame of the designated fast-forward frame number is calculated, and the calculated identification number is stored. By reading the video frame of the specified fast-forward frame number from the device, the same number of video frames can be read from each storage device in parallel, and as a result, efficient reading can be performed in proportion to the number of storage devices. Become. As a result, the number of playback frames per unit time at the time of fast-forward playback can be increased as compared with the conventional method, and as a result, smooth fast-forward or reverse fast-forward playback becomes possible. It is also possible to flexibly change the double speed value of fast forward or reverse fast forward by appropriately adjusting the frame number read from each storage device. As described above, according to the present invention, it is not necessary to separately prepare the data for normal reproduction and the data for fast-forward, and therefore, it is not necessary to perform control for transition between these two types of moving image data during fast-forward reproduction. 2. By making the number of storage devices m a prime number, a double speed coefficient n
Can be set to any integer. 3. By determining the number N of frames in the read unit such that mN is equal to the number of frames per second in full motion display, the method and apparatus of the present invention can be applied to a conventional moving picture program providing system. 4. By determining the remainder obtained by dividing the frame number of the moving image frame stored in each of the plurality of storage devices by m as the storage device number of the storage device, it becomes easy to calculate the storage device number during fast-forward playback. Also, after the storage device numbers are determined in this manner, the moving image frames are stored in each of the plurality of storage devices in the order of the storage device numbers in the order of the frame numbers. The moving image frames with the frame numbers can be stored in the same storage device in the order of the frame numbers.

[Brief description of the drawings]

FIG. 1 is a diagram schematically illustrating a system configuration according to an embodiment of the present invention.

FIG. 2 is a diagram schematically illustrating an accumulation state of moving image data accumulated in units of moving image frames according to an embodiment of the present invention.

FIG. 3 is a diagram showing operations at the time of normal reproduction and at the time of fast forward or reverse fast forward reproduction according to the embodiment of the present invention using a flowchart.

FIG. 4 is a diagram showing moving image data read by each magnetic disk during double-speed fast-forward reproduction according to the embodiment of the present invention.

FIG. 5 is a diagram showing moving image data read by each magnetic disk during triple-speed fast-forward reproduction in an embodiment of the present invention.

FIG. 6 is a diagram showing moving image data read by each magnetic disk during quadruple-speed fast-forward reproduction in an embodiment of the present invention.

FIG. 7 is a diagram schematically showing a system configuration in a conventional example.

[Explanation of symbols]

1 terminal 2 network 3 network interface unit 4 reads management unit 5 switch section 6 _0, 6 _1, 6 ₄ storage device controller 7 _0, 7 _1, 7 ₄ storage device 71 the terminal 72 network 73 network interface unit 74 switches

────────────────────────────────────────────────── (5) References JP-A-7-184164 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 5/76-5/956 G11B 20 / 10-20/12 H04N 9/79-9/898

Claims (8)

(57) [Claims] 1. A digital moving image storage / playback control method for storing digital moving images in a plurality of storage devices in frame units and playing back the playback in response to a playback request instructed from a terminal, wherein the number m of the plurality of storage devices is (M is an integer of 2 or more) , a digital moving image storage / playback control method for storing moving image frames having frame numbers that are the same as each other modulo m in the same storage device adjacent to each other in frame number order. 2. During normal reproduction, a predetermined number N of moving image frames stored in each storage device in the order of frame numbers are set as a reading unit, and moving image frames included in the reading unit are sequentially read and buffered. The ringing process is performed in parallel on the plurality of storage devices, and the buffered mN read from the plurality of storage devices is buffered.
The moving image frames are reproduced as a unit of reproduction, and the reading and buffering of the moving image frames and the reproduction of the buffered moving image frames are repeated in the order of the reading units of the moving image frames stored in each storage device. At the time of fast-forward playback, the playback frame number at the time of the fast-forward playback request is set as the first term, and the second or lower sequence of the arithmetic progression having a tolerance of a double speed coefficient n set to a relatively prime value with respect to m Specify the fast-forward frame number sequentially so that
Calculate the identification number of the storage device storing the video frame of the specified fast-forward frame number, read the video frame of the specified fast-forward frame number from the storage device of the calculated identification number, and buffer the video frame. 2. The digital moving image storage / playback control method according to claim 1, wherein, when the processing of reading and buffering the moving image frames makes a round of all storage devices, the moving image frames buffered are reproduced as a reproduction unit. 3. The method according to claim 2, wherein the number m of storage devices is a prime number. 4. The fast-forward playback includes a reverse fast-forward playback,
3. The method according to claim 2, wherein a double speed coefficient is set to a negative value when a reverse fast forward reproduction request is made. 5. The method according to claim 2, wherein a remainder obtained by dividing a frame number of a moving image frame stored in each of the plurality of storage devices by m is determined as a storage device number of the storage device. 6. A moving image frame is stored in the plurality of storage devices in the order of frame numbers in the order of the storage device numbers.
The method of claim 5. 7. The method according to claim 2, wherein the number N of frames in the readout unit is determined such that mN is equal to the number of frames per second in full motion display. 8. A digital moving image storage / playback control device for storing digital moving images in a storage device on a frame basis and playing back in response to a playback request, comprising: a plurality of storage devices arranged in parallel; A storage control device having a plurality of storage device control units having a buffer memory for controlling reading / writing and buffering moving image data read from the storage device; and a terminal and a storage control device via a network. An interface unit for interfacing between the plurality of storage devices, a switch unit for switching a connection between a data line for transmitting moving image data read in parallel from the plurality of storage devices and the interface unit, and a reproduction transmitted from the terminal. A request signal and a reproduction mode change request signal are received, and the storage controller reads the moving image data and reads the read moving image data. Instructs the buffering of the image data, and, and a readout control section for performing a reproduction instruction of the video data buffered, each storage device in said plurality of storage devices, the plurality of storage device number m ( m for an integer of 2 or more), each other modulo m
When the terminal receives a normal reproduction request from the terminal, the read management unit increments the reproduction frame number at that time by one, and increments the reproduction frame number by one. A read / buffer for calculating an identification number of a storage device storing a moving image frame having a selected frame number and sequentially reading and buffering a predetermined number N of moving image frames in order of a frame number starting from the moving image frame. The ring is instructed to the storage control device, and then the frame number is further incremented by one, and the identification number of the storage device storing the moving image frame of the incremented frame number is calculated. Instructs the storage controller to perform buffering, and
The buffering instruction is repeated until all the storage devices are cycled, and the data is read out in parallel from each storage device in response to the read / buffering instruction executed in a cycle through all the storage devices. , And instructs the accumulation control device and the switch unit to perform normal reproduction using mN moving image frames as a reproduction unit. When a fast-forward reproduction request is issued from the terminal, the read management unit sets a reproduction frame number at that time to m. Increments by a double speed coefficient n set to a relatively prime value,
The identification number of the storage device that stores the incremented moving image frame of the frame number is calculated, and the storage control device instructs the storage control device to read and buffer the moving image frame from the storage device of the calculated identification number. In addition, the frame number is further incremented by n, the identification number of the storage device that stores the moving image frame of the incremented frame number is calculated, and reading and buffering of the moving image frame are instructed to the storage control device, This read and buffering instruction is repeated until all storage devices have been cycled, and read in parallel from each of the storage devices in response to the read and buffering instructions executed for all of the storage devices. The fast-forward playback using the output m movie frames as a playback unit is performed by the storage control device and the switch unit. The storage control device receives the normal reproduction instruction or the fast-forward reproduction instruction from the read management unit, and repeats the read and buffering instruction and the fast-forward reproduction instruction. The data is read out onto a data line between the storage control unit and the switch unit, and the switch unit interfaces with the storage control unit under the control of the read management unit to transfer the read moving image frame to the terminal. A digital moving image storage / playback control device, wherein a signal path between ace sections is set.