JP2023064284A - Communication device, communication method, and program - Google Patents

Abstract

To prevent deterioration in measurement accuracy of a residence time when a plurality of independent synchronous paths fail.SOLUTION: A communication device according to one aspect includes: a first communication interface used for relaying a time synchronization packet on a first route; a second communication interface used for relaying a time synchronization packet on a second route; first clocking means provided in the first communication interface; second clocking means provided in the second communication interface; and synchronization means for synchronizing the first clocking means and the second clocking means based on the link state of at least any one of the first route and the second route.SELECTED DRAWING: Figure 1

Description

The present invention relates to a communication device, communication method and program.

Attention has been paid to a technique of installing a plurality of cameras at different positions to perform synchronous imaging from multiple viewpoints, and generating virtual viewpoint content using the multi-viewpoint images obtained by the synchronous imaging. Virtual viewpoint technology makes it possible to view the highlight scenes of sports and other events from various angles, providing viewers with a high sense of presence.

Patent Document 1 discloses a method of extracting image data of a predetermined region of images captured by a plurality of cameras and generating a virtual viewpoint image using the extracted image data. The image processing devices are connected in a daisy chain, and image data from each image processing device is transmitted to the image generation device through the daisy chain network. Here, each control unit has a function of PTPv2 (Precision Time Protocol version 2) of the IEEE 1588 standard, and synchronizes the imaging timing of a plurality of cameras by performing processing related to time synchronization with a time server.

In the following description, a terminal having a clock that serves as a reference for time synchronization is sometimes called a synchronization master, and a terminal that adjusts to the reference clock for time synchronization is sometimes called a synchronization slave.

Patent Document 2 describes a method of correcting the time error that statically occurs in each path and changing the path of the time synchronization packet when a problem occurs in one path in a BC terminal that has two synchronous paths. is disclosed.

In the time synchronization, a time stamp mechanism provided in the I/F (Interface) is used to measure the time (dwell time) required for relaying the time synchronization packet.

JP 2017-211828 A JP 2019-158538 A

However, when a time synchronization server is provided for each of the two independent synchronization paths, if one of the two independent synchronization paths is disconnected, the time synchronization packet may be detoured to the undisconnected path. . In this case, the time stamp of a clock synchronized with a different time synchronization server is used in the route used for detouring the time synchronization packet, and the measurement accuracy of the retention time may be lowered.
The problem to be solved by the present invention is to suppress the deterioration of the measurement accuracy of the residence time when a plurality of independent synchronous paths fail.

The communication device according to one aspect includes a first communication interface used for relaying the time synchronization packet on the first path, a second communication interface used for relaying the time synchronization packet on the second path, and the second Based on the link state of at least one of the first timekeeping means provided in one communication interface, the second timekeeping means provided in the second communication interface, and the first path and the second path, the Synchronizing means for synchronizing the first timekeeping means and the second timekeeping means.

ADVANTAGE OF THE INVENTION According to one aspect of the present invention, it is possible to suppress a decrease in the measurement accuracy of the residence time when a plurality of independent synchronous paths fail.

1 is a block diagram showing a configuration example of a synchronization system according to a first embodiment; FIG. FIG. 2 is a block diagram showing a configuration example of a synchronized imaging system to which the synchronization system of FIG. 1 is applied; FIG. 3 is a block diagram showing a configuration example of a camera adapter applied to the synchronous imaging system of FIG. 2; FIG. 2 is a block diagram showing the flow of detouring time synchronization packets when a link down occurs in the synchronization system of FIG. 1; FIG. 2 is a block diagram showing the flow of time synchronization packets without detouring when a link down occurs in the synchronization system of FIG. 1; FIG. 2 is a diagram showing relay destinations of time synchronization packets when a link down occurs in the synchronization system of FIG. 1; FIG. 2 is a diagram showing a method of synchronizing a clock unit when no link down occurs in the synchronous system of FIG. 1; FIG. 2 is a diagram showing an example of a synchronization method of a clock section when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; FIG. 10 is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1; 4 is a block diagram showing the operation of the clock synchronization switching unit in FIG. 3; FIG. FIG. 4 is a diagram showing a synchronization source of a switch and a clock unit for each link state in the clock synchronization switching unit of FIG. 3; FIG. 3 is a diagram showing a sequence of time synchronization of the synchronous imaging system of FIG. 2; 4 is a flowchart showing the operation of the camera adapter of FIG. 3; FIG. 11 is a diagram showing synchronization sources of switches and a clock unit according to link status of the clock synchronization switching unit according to the second embodiment; 13 is a flow chart showing the operation of the clock synchronizer according to the fourth embodiment;

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be noted that the following embodiments do not limit the present invention, and not all combinations of features described in the embodiments are essential for solving means of the present invention. The configuration of the embodiment can be appropriately modified or changed according to the specifications of the device to which the present invention is applied and various conditions (use conditions, use environment, etc.). The technical scope of the present invention is defined by the claims and is not limited by the following individual embodiments.

<First Embodiment>
FIG. 1 is a block diagram showing a configuration example of a synchronization system according to the first embodiment.
In FIG. 1, the synchronization system 100 comprises two independent synchronization paths A,B. The synchronization system 100 includes time servers 104-A, 104-B, a hub 105 and TC (Transparent Clock) terminals 101-a to 101-e. Each of the time servers 104-A and 104-B generates a reference time for time synchronization for each of the synchronization paths A and B. Each time server 104-A, 104-B can be used as a time synchronization server. Each time server 104-A, 104-B can act as a synchronization master. The hub 105 relays the time synchronization packets transmitted from the time servers 104-A and 104-B to the TC terminal 101-a.

For example, a PTP packet can be used as the time synchronization packet. PTP (Precision Time Protocol) defines BC (Boundary Clock) and TC (Transparent Clock), and defines methods for establishing and operating a synchronization network. A BC synchronizes with a synchronization master such as a GMC (Grand Master Clock), and further operates as a synchronization master itself to separate and extend a synchronization network. When the TC receives a PTP packet from the GMC or synchronization master, the TC calculates the retention time of the time synchronization packet in the TC when transferring it to the synchronization slave downstream of the network, and gives it to the PTP packet as a correction value. The synchronization slave can achieve highly accurate time synchronization by using the dwell time assigned to the PTP packet. The TC does not need to be time-synchronized with the GMC and the synchronization master, but time-synchronization can improve the synchronization accuracy of the entire system.

Here, each of the TC terminals 101-a to 101-e is provided with two independent synchronization paths A and B for transferring time synchronization packets and connected in a daisy chain. Each of the TC terminals 101-a to 101-e includes four communication I/F units (communication interfaces) 102-a to 102-d, and each communication I/F unit 102-a to 102-d includes a clock 103-a to 103-d.

On route A, a time synchronization packet transmitted from time server 104-A is received by communication I/F section 102-a of TC terminal 101-a via hub 105. FIG. The time synchronization packet received by the communication I/F unit 102-a of the TC terminal 101-a is transferred to the communication I/F unit 102-c and sequentially transmitted to the subsequent TC terminals 101-b to 101-e. be.

On route B, the time synchronization packet transmitted from time server 104-B is received by communication I/F section 102-b of TC terminal 101-a via hub 105. FIG. The time synchronization packet received by the communication I/F unit 102-b of the TC terminal 101-a is transferred to the communication I/F unit 102-d and sequentially transmitted to the subsequent TC terminals 101-b to 101-e. be.

Each of the communication I/F units 102-a to 102-d time-synchronized packets passing through the routes A and B, clock units 103-a to 103-d of each of the communication I/F units 102-a to 102-d Measure the time (dwell time) required for relaying in the TC terminals 101-b to 101-e. Specifically, each of the communication I/F units 102-a and 102-b uses the clock units 103-a and 103-b of the communication I/F units 102-a and 102-b when receiving the time synchronization packet. Measure the reception time. Then, each communication I/F unit 102-c, 102-d uses the clock unit 103-c, 103-d of each communication I/F unit 102-c, 102-d to set the transmission time when transmitting the time synchronization packet. measure.

Further, each of the TC terminals 101-a to 101-e not only relays the time synchronization packet, but also operates as a synchronization slave to perform time synchronization with the time servers 104-A and 104-B. Specifically, each of the TC terminals 101-a to 101-e synchronizes the clock section 103-a of the communication I/F section 102-a with the time synchronization packet transmitted by the time server 104-A. Further, each of the TC terminals 101-a to 101-e synchronizes the clock section 103-b of the communication I/F section 102-b with the time synchronization packet transmitted by the time server 104-B. Further, each of the TC terminals 101-a to 101-e internally synchronizes the clock units 103-a and 103-c, and synchronizes the clock units 103-b and 103-d.

As described above, in the synchronization system 100, the clock units 103-a and 103-c are synchronized on the route A, and the clock units 103-b and 103-c are synchronized on the route B in parallel with the route A. is done. Therefore, when an abnormality occurs in one path, it is possible to immediately switch to the clock of the other path, and the synchronization system 100 can maintain time synchronization without lowering the synchronization accuracy.

FIG. 2 is a block diagram showing a configuration example of a synchronized imaging system to which the synchronization system of FIG. 1 is applied. In the synchronous imaging system 200 of FIG. 2, a plurality of cameras 220-a to 220-z are installed in a facility such as a stadium, a concert hall, or a studio, and synchronous imaging is performed.

In FIG. 2, synchronized imaging system 200 comprises image sensor systems 290 - a through 290 - z, image computing server 260 , user terminal 270 , control terminal 280 , time servers 104 -A, 104 -B and hub 105 . Each image sensor system 290-a through 290-z comprises a camera 220-a through 220-z and a camera adapter 201-a through 201-z.

Each image sensor system 290-a through 290-z captures synchronized motion images and synchronized still images and transmits them to image computing server 260 over a daisy-chained network. At this time, each of the image sensor systems 290-a to 290-z transmits and receives imaging data, control data, and time synchronization packets via the daisy chained camera adapters 201-a to 201-z and the hub 105. conduct.

Each camera 220-a through 220-z captures an image. Each camera 220-a to 220-z has an optical system including a focus lens for focusing on the subject and a zoom lens for adjusting the angle of view, and an imaging device that converts the light from the subject into an electric signal for each pixel. Prepare.

Each camera adapter 201-a through 201-z captures captured images captured by respective cameras 220-a through 220-z and transmits them to image computing server 260. FIG. Each camera adapter 201-a to 201-z includes the functions of the TC terminal shown in FIG. 1 and performs time synchronization.

Image computing server 260 processes the synchronized images received from each image sensor system 290-a through 290-z. The image computing server 260 then generates virtual viewpoint content using the multi-viewpoint images obtained by the synchronized imaging.
The user terminal 270 displays the virtual viewpoint content generated by the image computing server 260 and instructs the viewing angle of the virtual viewpoint content.
The control terminal 280 manages the operation state and controls parameter setting, etc. for each device constituting the synchronized imaging system 200 through the network.

The image sensor systems 290-a to 290-z are connected by a daisy chain and the connection is duplicated. At this time, image sensor systems 290-a through 290-z include a path A daisy chain 210-aa through 210-az and a path B daisy chain 210-ba through 210-bz. By duplicating the daisy chain connection, it is possible to maintain the time synchronization accuracy of the synchronized imaging system 200 even if a failure occurs in one of the connection systems.

The daisy chains 210-aa to 210-az, 210-ba to bz are, for example, Ethernet (registered trademark) GbE (Gigabit Ethernet) conforming to the IEEE standard, or 10GbE or 100GbE. However, it is not limited to this, and other types of networks may be used. The lines connecting hub 105 and time servers 104-A and 104-B, hub 105 and image computing server 260, hub 105 and control terminal 280, and hub 105 and user terminal 270 are similar.

The operation of the synchronized imaging system 200 will be described below.
An image captured by camera 220-z is transmitted by camera adapter 101-z to camera adapter 101-y of image sensor system 290-y using one of daisy chains 210-az and 210-bz. Image sensor system 290-y transmits an image captured by camera 220-y together with an image acquired from image sensor system 290-z to an adjacent image sensor system.

By continuing such processing, images acquired by each of image sensor systems 290-a to 290-z are transferred from image sensor system 290-a using one of daisy chains 210-aa and 210-ba. sent to hub 105 . Images sent to hub 105 are transmitted to image computing server 260 .

Image computing server 260 reconstructs and converts the data format of image packets obtained from image sensor system 290-a, and then stores them according to the camera identifier, data type and frame number. Then, the image computing server 260 accepts the designation of a viewpoint from the control terminal 280, reads the corresponding image data from the stored data based on the accepted viewpoint, and performs rendering processing to generate a virtual viewpoint image. That is, the image computing server 260 generates virtual viewpoint content based on captured images (multi-viewpoint images) captured by the plurality of cameras 220-a to 220-z and viewpoint information.

The rendered image is transmitted from the image computing server 260 to the user terminal 270, and the user operating the user terminal 270 can browse the image of the designated viewpoint.

On the other hand, each of the time servers 104-A and 104-B transmits a time synchronization packet to each of the duplicated daisy chains 210-aa to 210-az and 210-ba to bz. Each image sensor system 290-a to 290-z receives the time synchronization packet on duplicated daisy chains 210-aa to 210-az and 210-ba to bz, respectively.

Then, each image sensor system 290-a to 290-z uses the TC function inside each camera adapter 201-a to 201-z, based on the time synchronization packet received by the communication I/F unit 102-a, the clock unit 103 - Perform time synchronization of a.

Similarly, the image sensor systems 290-a to 290-z perform time synchronization of the clock section 103-b based on the time synchronization packet received by the communication I/F section 102-b. Then, the image sensor systems 290-a to 290-z generate Genlock signals based on the time synchronization of the clock units 103-a and 103-b, and perform image capturing synchronization.

In addition, camera adapter 201-a receives the time synchronization packet transmitted by time server 104-A through Ethernet 210-aa at communication I/F unit 102-a, synchronizes clock unit 103-a, and communicates I /F section 102-c.

Similarly, the camera adapter 201-a receives the time synchronization packet transmitted by the time server 104-B through the Ethernet 210-ba at the communication I/F unit 102-b, synchronizes the clock unit 103-b, and communicates It is relayed from the I/F section 102-d.
By performing such operations by the camera adapters 201-a to 201-y, image capturing synchronization of the image sensor systems 290-a to 290-z can be performed.

In this way, all the camera adapters 201-a to 201-z of the synchronized imaging system 200 are provided with a function of time synchronization and a function of transferring time synchronization packets distributed by the time servers 104-A and 104-B. , the accuracy of time synchronization can be improved. The camera adapters 201-a to 201-z, which are time-synchronized based on the times of the time servers 104-A and 104-B, can synchronize the imaging timings of the cameras 220-a to 220-z. becomes. As a result, the synchronized imaging system 200 can generate a virtual viewpoint image based on a plurality of captured images captured at the same timing, and can suppress deterioration in the quality of the virtual viewpoint image due to a shift in imaging timing.

FIG. 3 is a block diagram showing a configuration example of a camera adapter applied to the synchronized imaging system of FIG. Note that the camera adapter 201 in FIG. 3 can be used as each of the camera adapters 201-a to 201-z in FIG.

Of the functional blocks shown in FIG. 3, for functions realized by software, a program for providing the function of each functional block is stored in a memory such as a ROM (Read Only Memory). Then, the program is read into a RAM (Random Access Memory) and executed by a CPU (Central Processing Unit). For the functions realized by hardware, for example, by using a predetermined compiler, a dedicated circuit may be automatically generated on the FPGA from a program for realizing the function of each functional block. FPGA is an abbreviation for Field Programmable Gate Array. Also, a gate array circuit may be formed in the same manner as the FPGA and implemented as hardware. Also, it may be realized by an ASIC (Application Specific Integrated Circuit). Note that the configuration of the functional blocks shown in FIG. 3 is an example, and a plurality of functional blocks may constitute one functional block, or one of the functional blocks may be divided into blocks that perform a plurality of functions. good too.

3, the camera adapter 201 includes communication I/F units 102-a to 102-d, clock synchronization units 301-a and 301-b, abnormality recovery detection units 303-a to 303-d, time synchronization control unit 304 -a to 304-d and path switching units 305-a to 305-d. The camera adapter 201 also includes clock synchronization switching units 306 - a and 306 - b, an image processing unit 307 and a camera control unit 308 . Each of communication I/F units 102-a to 102-d includes clock units 103-a to 103-d. The image processing unit 307 and camera control unit 308 are connected to the camera 220 .

Each clock unit 103-a to 103-d keeps time. Each clock unit 103-a to 103-d is a hardware clock that holds the current time. Each of the clock units 103-a to 103-d counts up based on a clock signal output by a device such as crystal, and periodically outputs a reference signal that serves as a reference for the time within the camera adapter 201. FIG. Each of the clock units 103-a to 103-d adjusts the frequency by increasing or decreasing the counted-up value based on the instruction from the time synchronization control unit 304, and changes the progress of time.

A device such as a crystal always outputs a clock signal at a unique frequency, but there are individual differences, and the frequency changes due to environmental changes such as temperature conditions or vibration. Therefore, there is a discrepancy between the progress of the individual clocks of the clock units 103-a to 103-d and the progress of the clocks of the time servers 104-A and 104-B. In order to correct this deviation, the frequency of each clock unit 103-a to 103-d is adjusted and controlled to synchronize with each time server 104-A, 104-B. Each of the clock units 103-a to 103-d can set not only synchronization control by frequency but also time itself. If the time difference with each time server 104-A, 104-B is large at the start of time synchronization, set the time of each clock unit 103-a to 103-d and use it to adjust the frequency. can be done.

By connecting two of the four communication I/F units 102-a to 102-d of the camera adapter 201 to adjacent camera adapters, a double daisy chain can be configured. However, the two communication I/F units 102-a and 102-b of the first-stage camera adapter 201-a in FIG.

Each of the clock units 103-a to 103-d has a time stamp function that stores the transmission time and the reception time when time synchronization packets are transmitted and received. Note that each of the clock units 103-a to 103-d may have a buffer such as a FIFO (First In First Out) so that the time stamps of a plurality of packets can be saved.
Each of the time synchronization control units 304-a to 304-d acquires the received time synchronization packet and the value of the time stamp, and uses the contents of the time of the time synchronization packet and the value of the time stamp to each clock unit 103-a to The frequency of 103-d is adjusted to perform time synchronization.

An image captured by the camera 220 is separated into a foreground image and a background image by the image processing unit 307, and the image is captured via the camera adapter 201 and the hub 105 using one of the communication I/F units 102-a and 102-b. It is sent to computing server 260 . By outputting a foreground image and a background image from each camera adapter 201, a virtual viewpoint image is generated based on the foreground image and the background image captured from a plurality of viewpoints. Note that there may be a camera adapter 201 that outputs the foreground image separated from the captured image and does not output the background image.

A captured image transmitted from camera adapter 201-b in FIG. output from one of a and 102-b. The communication I/F units 102-a and 102-b used for output are switched according to the state of the camera adapter 201. FIG.

Each of the abnormality recovery detection units 303-a to 303-d detects the link down of each communication I/F unit 102-a to 102-d and the time-out of the time synchronization packet, and detects the restoration thereof. At this time, the error recovery detection unit 303-a detects an error and recovery occurring in the communication unit I/F 102-a and the synchronization control unit 304-a. The error recovery detection unit 303-b detects an error and recovery occurring in the communication unit I/F 102-b and the synchronization control unit 304-b. The error recovery detection unit 303-c detects an error and recovery occurring in the communication I/F unit 102-c. The error recovery detection unit 303-d detects an error and recovery occurring in the communication I/F unit 102-d.

The information of the abnormality detection and restoration detection detected by each of the abnormality restoration detection units 303-a to 303-d is the time synchronization control units 304-a, 304-b, the clock synchronization switching units 306-a, 306-b, and the route It is transmitted to the switching units 305-a to 305-d. Incidentally, each of the abnormality recovery detection units 303-a to 303-d may be included in the communication I/F units 102-a to 102-d, and each of the abnormality recovery detection units 303-a and 303-b It may be included in each time synchronization control unit 304-a, 304-b.

Each of the time synchronization control units 304-a and 304-b performs protocol processing for time synchronization by a method conforming to the IEEE1588-2008 standard. Camera adapter 201 transmits and receives time synchronization packets to and from time servers 104-A and 104-B. Camera adapter 201 then synchronizes communication I/F units 102-a to 102-d with either time server 104-A or 104-B, and relays time synchronization packets.

In addition, each time synchronization control unit 304-a, 304-b, the time of the time server 104-A, 104-B obtained from the packet, and the time stamp obtained from each communication I / F unit 102-a ~ 102-d Synchronize based on time difference. At this time, each of the time synchronization control units 304-a and 304-b sets the time of the clock units 103-a and 103-d and adjusts the frequency for synchronization. For example, the time synchronization control unit 304-a synchronizes the clock unit 103-a linked to the communication I/F unit 102-a with the time server 104-A. The time synchronization control unit 304-b synchronizes the clock unit 103-b linked to the communication I/F unit 102-b with the time server 104-B.

Furthermore, each of the time synchronization control units 304-a and 304-b measures the retention time during which the time synchronization packet stays within its own terminal, and adds the retention time to a specific area of the communication packet to be transferred. Each time synchronization control unit 304-a, 304-b receives the time stamp of the communication I / F unit 102-a, 102-b at the time of receiving the time synchronization packet, and the communication I / The retention time is obtained from the transmission time stamps of the F units 102-c and 102-d. Each time synchronization control unit 304-a, 304-b adds the dwell time to a specific area of the communication packet, so that the time for the communication packet to pass through each TC terminal 101-a to 101-e is also time-controlled. can be taken in. Therefore, each of the time synchronization control units 304-a and 304-b can improve the accuracy of frequency correction for time synchronization.

Note that IEEE1588-2008 has the concept of PTP domain. In the PTP domain, synchronization with the time synchronization packet can be performed only when the domain of the time synchronization packet transmitted by each of the time servers 104-A and 104-B is the same as the domain specified for the client performing time synchronization. In the example of FIG. 2, the PTP domain of the time server 104-A and the time synchronization controller 304-a and the PTP domain of the time server 104-B and the time synchronization controller 304-b can be configured as different domains. This makes it possible to increase the independence of the network of the daisy chains 210aa-210az and the network of the daisy chains 210ba-210bz.

Each of the clock synchronization units 301-a and 301-b synchronizes the clock units 103-a to 103-d. For example, the clock synchronization unit 301-a uses the clock of the clock unit 103-a as a synchronization source and synchronizes the clock unit 103-c, or uses the clock of the clock unit 103-b as a synchronization source and synchronizes the clock unit 103-c. Synchronize. The clock synchronization unit 301-b synchronizes the clock unit 103-d with the clock of the clock unit 103-b as the synchronization source, or synchronizes the clock unit 103-d with the clock of the clock unit 103-a as the synchronization source. Synchronize. At this time, each of the clock synchronization units 301-a and 301-b may gradually synchronize the clock units 103-a to 103-d.

Each of the clock synchronization switching units 306-a and 306-b switches the clock units 103-a to 103-d with which the clock synchronization units 301-a and 301-b are synchronized. For example, the clock synchronization switching unit 306-a can synchronize the time of the clock unit 103-a with the time of the clock unit 103-c. In addition, the clock synchronization switching unit 306-a synchronizes the time of the clock unit 103-a with the time of the clock unit 103-c and the time of the clock unit 103-d. can synchronize the time of the clock units 103-c and 103-d.

Each path switching unit 305-a to 305-d receives the time synchronization packet received by each communication I/F unit 102-a, 102-b from either communication I/F unit 102-c, 102-d Reroute time synchronization packets to be sent.

FIG. 4 is a block diagram showing the flow of detouring time synchronization packets when a link down occurs in the synchronization system of FIG. The white arrow indicates the time synchronization packet sent by the time server A 104-A, and the black arrow indicates the time synchronization packet sent by the time server B 104-B. Also, a, b, c, and d indicate the positions of communication I/F units 102-a to 102-d, respectively.

In FIG. 4, the camera adapter 201-a is in a state where no abnormality is detected by the abnormality recovery detection units 303-a to 303-d in FIG. At this time, the route switching units 305-a and 305-c change the route of the time synchronization packet so that the time synchronization packet received by the communication I/F unit 102-a is transmitted from the communication I/F unit 102-c. switch. Path switching units 305-b and 305-d switch the path of the time synchronization packet so that the time synchronization packet received by communication I/F unit 102-b is transmitted from communication I/F unit 104-d. As a result, the time synchronization packet received by the communication I / F unit 102-a is transmitted from the communication I / F unit 102-c, and the time synchronization packet received by the communication I / F unit 102-b is the communication I /F unit 102-d.

Further, the camera adapter 201-b is in a state where the link down of the communication I/F section 102-c has been detected by the abnormality recovery detection section 303-c. At this time, the time synchronization packet transmitted by the time server 104-A is received by the communication I/F unit 102-a of the camera adapter 201-b, but the communication I/F unit 102-c is linked down. . In this case, path switching section 305-a switches the relay destination of the time synchronization packet received by communication I/F section 102-a to communication I/F section 102-d. On the other hand, the path switching unit 305-b keeps the communication I/F unit 102-d as the relay destination for the time synchronization packet transmitted by the time server 104-B. As a result, both the time synchronization packet transmitted by time server 104-A and the time synchronization packet transmitted by time server 104-B are relayed to communication I/F section 102-d.

Further, the camera adapter 201-c is in a state where the link down of the communication I/F unit 102-a has been detected by the abnormality recovery detection unit 303-a. In this case, since the communication I/F section 102-a is linked down, the time synchronization packet is not received. Therefore, the path switching unit 305-a does not operate. On the other hand, communication I/F section 102-b receives the packet of time server 104-A and the packet of time server 104-B. Therefore, the path switching unit 305-b determines whether the time synchronization packet is a packet for the time server 104-A or a packet for the time server 104-B. Then, the route switching unit 305-b switches the route to the communication I / F unit 102-d if it is a packet of the time server 104-B, and if it is a packet of the time server 104-A, the communication I / F unit Switch the route to 102-c. The path switching unit 305-b can use the PTP domain to determine whether the packet is transmitted by the time server 104-A or the time server 104-B.

Further, the camera adapter 201-d is in a state where the link down of the communication I/F section 102-d is detected by the abnormality recovery detection section 303-d. In this case, path switching units 305-b and 305-c switch the relay destination to communication I/F unit 102-c in the same manner as when the link down of communication I/F unit 102-c is detected.

Further, the camera adapter 201-e is in a state where the link down of the communication I/F section 102-b has been detected by the abnormality recovery detection section 303-b. In this case, the path switching unit 305-a determines whether the time server 104-A or 104-B and switches the relay destination, as in the case where the link down of the communication I/F unit 102-a is detected.

FIG. 5 is a block diagram showing the flow of time synchronization packets when a link down occurs in the synchronization system of FIG.
Link down of path A between TC terminal 101-b and TC terminal 101-c in FIG. -a link down) has occurred. Furthermore, the link down of path B between TC terminal 101-d and TC terminal 101-e (communication I/F section 102-d of TC terminal 101-d and communication I/F section 102- of TC terminal 101-e It is assumed that the link down of b) has occurred. In this case, the two routes A and B after the TC terminal 101-e (after the TC terminal 101-f not shown) become disconnected, and the time synchronization packet does not arrive, so high-precision synchronization is maintained throughout the system. I can't.

As shown in FIG. 5, when an abnormality is detected by the abnormality detection/restoration unit 303-a of the camera adapter 201-c, the time synchronization packet transmitted from the time server 104-A is transmitted to the communication I of the camera adapter 201-c. It cannot be received by the /F unit 102-a. Therefore, the camera adapter 201-c cannot relay the time synchronization packet transmitted from the time server 104-A.

Further, when an abnormality is detected by the abnormality detection/recovery unit 303-b of the camera adapter 201-e, the time synchronization packet transmitted from the time server 104-B is transmitted to the communication I/F unit 102- of the camera adapter 201-e. b cannot receive. Therefore, the camera adapter 201-e cannot relay either the time synchronization packet transmitted from the time server 104-A or the time synchronization packet transmitted from the time server 104-B. As a result, the camera adapters after the camera adapter 201-f cannot perform time synchronization.

When such a link down occurs, as shown in FIG. 4, by switching the communication I/F units 102-a to 102-d of the relay destination of the time synchronization packet, the camera adapter after the camera adapter 201-e time synchronization can be performed. At this time, each of the camera adapters 201-a to 201-e adds the retention time required for relaying the packet to a specific area of the time synchronization packet, and uses the value for time correction so that the time server 104-A , 104-B.

FIG. 6 is a diagram showing relay destinations of time synchronization packets when a link down occurs in the synchronization system of FIG.
In addition, in FIG. 6, the communication I / F unit that has been linked down, the communication I / F unit that received the time synchronization packet, the time server that transmitted the time synchronization packet, and the communication I that relays the time synchronization packet received from these conditions /F indicates which part will be. 6 a to d are communication I / F units 102-a to 102d, "recv PTP-A" receives the time synchronization packet PTP-A transmitted by the time server 104-A, and "recv PTP-B" is It indicates that the time synchronization packet PTP-B transmitted by the time server 104-B is received.

In FIG. 6, for example, when the communication I / F unit 102-d is linked down and the time synchronization packet of the time server 104-B is received by the communication I / F unit 102-b, the relay destination of the time synchronization packet is It becomes the communication I/F unit 102-c. Further, when the communication I / F unit 102-b is linked down, when the time synchronization packet of the time server 104-B is received by the communication I / F unit 102-a, the relay destination of the time synchronization packet is the communication I / F It becomes part 102-d.

In one camera adapter 201, if communication I/F units 102-a and 102-b are linked down at the same time, the time synchronization packet cannot be received and therefore the time synchronization packet cannot be relayed. The same applies when three or more communication I/F units 102-a to 102-d in one camera adapter 201 are linked down at the same time. Also, if the communication I/F units 102-c and 102-d are linked down at the same time, the time synchronization packet cannot be relayed.

FIG. 7A is a diagram showing a method of synchronizing the clock unit when link down does not occur in the synchronization system of FIG.
In FIG. 7A, the white arrow indicates the time synchronization packet sent by the time server 104-A, and the black arrow indicates the time synchronization packet sent by the time server 104-B. Clock units 103-a to 103-d are provided in communication I/F units 102-a to 102-d.

Circle clocks in communication I/F units 102-a and 02-c in FIG. - Indicates that it is in sync with B. Also, when the communication I/F units 102-a to 102-d are blank, it indicates that they are not synchronized with any of the time servers 104-A and 104-B. Also, the arrows connecting the clock units 103-a to 103-d, the clock synchronization switching units 306-a and 306-b, and the clock synchronization units 301-a and 301-b in FIG. 7A indicate the directions for adjusting the time. . The same applies to FIGS. 7B-7K.

In FIG. 7A, the daisy-chained routes A and B are independent of each other, the time synchronization packet PKT A transmitted by the time server 104-A passes through the route A, and the time synchronization packet PKT B transmitted by the time server 104-B passes through path B.

The time synchronization control unit 304-a in FIG. 3 uses the time synchronization packet PKT A transmitted by the time server 104-A to change the progress of the time of the clock unit 103-a to change the time of the clock unit 103-a. Synchronize. The time synchronization control unit 304-b uses the time synchronization packet PKT B transmitted by the time server 104-B to change the progress of the time of the clock unit 103-b to synchronize the time of the clock unit 103-b. .

Assume that the communication I/F units 102-a to 102-d of the camera adapter 201 are normal. In this case, the clock synchronization switching unit 306-a connects the clock unit 103-a and the clock synchronization unit 301-a so that the clock synchronization unit 301-a refers to the clock unit 103-a as the synchronization source clock. Thereby, the clock synchronization section 301-a synchronizes the clock section 103-c with the clock section 103-a. At this time, clock unit 103-c indirectly synchronizes with time server 104-A.

Further, the clock synchronization switching unit 306-b connects the clock unit 103-b and the clock unit clock synchronization unit 301-b so that the clock synchronization unit 301-b refers to the clock unit 103-b as the synchronization source clock. do. Thereby, the clock synchronization section 301-b synchronizes the clock section 103-d with the clock section 103-b. At this time, clock unit 103-d indirectly synchronizes with time server 104-B.

Clock unit 103-c is synchronized with clock unit 103-a, so that clock unit 103-a of communication I/F unit 102-a that has received time synchronization packet PKT A and communication I/F unit 102- of the relay destination The clock section 103-c of c can be synchronized. Clock unit 103-d synchronizes with clock unit 103-b, so that clock unit 103-b of communication I/F unit 102-b that receives time synchronization packet PKT B and communication I/F unit 102- of the relay destination d can be synchronized.

A case where one of the communication I/F units 102-a to 102-d is linked down will be described below.
7B is a diagram showing an example of a method of synchronizing the clock unit when a link down occurs in the synchronization system of FIG. 1. FIG.
In FIG. 7B, it is assumed that the abnormality restoration detection unit 303-a of FIG. 3 has detected a link down (abnormality) of the communication I/F unit 102-a. When the communication I/F unit 102-a is linked down, if the connection of the clock synchronization switching unit 306-a is maintained as shown in FIG. 7A, the clock unit 103-c will synchronize with the time server 104-A. Synchronize with clock unit 103-a that is not present. In this case, the camera adapter 201 cannot correctly measure the retention time when relaying the time synchronization packet PKT A of the time server 104-A received by the communication I/F unit 102-b from the communication I/F unit 102-c. .

Therefore, clock synchronization switching section 306-b sets clock synchronization sections 301-a and 301-b so that clock synchronization sections 301-a and 301-b refer to clock section 103-b as a synchronization source clock. -b. By this connection, the clock units 103-c and 103-d are synchronized with the clock unit 301-b. This makes it possible to accurately measure the retention time when relaying the time synchronization packet PKT A transmitted from the time server 104-A from the communication I/F unit 102-b to the communication I/F unit 102-c. can. In addition, it is possible to accurately measure the retention time when relaying the time synchronization packet PKT B transmitted from the time server 104-B from the communication I / F unit 102-b to the communication I / F unit 102-d. .

FIG. 7C is a diagram showing another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1;
In FIG. 7C, it is assumed that communication I/F section 102-b is linked down. In this case, the clock synchronization switching unit 306-a synchronizes the clock unit 103-a with the clock synchronization unit 301-a, so that the clock synchronization units 301-a and 301-b refer to the clock unit 103-a as the synchronization source clock. 301-b.

FIG. 7D is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1;
In FIG. 7D, it is assumed that communication I/F section 102-c is linked down. In this case, in order to measure the retention time when relaying the time synchronization packet PKT A from the time server 104-A received by the communication I / F unit 102-a to the communication I / F unit 102-d, the clock unit The I/F 103-a is synchronized with the clock I/F 103-d. In addition, in order to measure the retention time when relaying the time synchronization packet PKT B from the time server 104-B received by the communication I / F unit 102-b to the communication I / F unit 102-d, the clock unit I /F 103-b is synchronized with the clock unit 103-d. At this time, the time synchronization control unit 304-a stops the synchronization between the time server 104-A and the clock unit 103-a even though time synchronization using the time server 104-A and the time synchronization packet PKT A is possible. do.

Then, clock synchronization switching section 306-b sets clock synchronization sections 301-a and 301-b so that clock synchronization sections 301-a and 301-b refer to clock section 103-b as a synchronization source clock. -b. Further, the clock synchronization switching unit 306-a connects the clock synchronization unit 301-a to the clock unit 103-a so that the clock unit 103-a is synchronized with the clock synchronization unit 301-a. By this connection, the clock units 103-a, 103-c, and 103-d are synchronized with the clock unit 103-b. Thereby, when the communication I / F unit 102-c is linked down, the time synchronization packet PKT A transmitted from the time server 104-A from the communication I / F unit 102-a to the communication I / F unit 102-d It is possible to accurately measure the residence time when relaying. In addition, it is possible to accurately measure the retention time when relaying the time synchronization packet PKT B transmitted from the time server 104-B from the communication I / F unit 102-b to the communication I / F unit 102-d. .

FIG. 7E is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1;
In FIG. 7E, it is assumed that communication I/F section 102-d is linked down. In this case, the time synchronization control unit 304-b stops synchronizing the time server 104-B and the clock unit 103-b. The clock synchronization switching section 306-a connects the clock synchronization sections 301-a and 301-b so as to synchronize with the clock section 103-a. Also, the clock synchronization switching unit 306-b connects the clock unit 103-b to synchronize with the clock synchronization unit 301-b.

A case where two of the communication I/F units 102-a to 102-d are linked down will be described below.
FIG. 7F is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1;
In FIG. 7F, it is assumed that communication I/F units 102-a and 102-c on route A are linked down. In this case, the clock synchronization switching section 306-b connects the clock section 103-b to the clock synchronization section 301-b and synchronizes the clock section 103-d with the clock section 103-b. As a result, even when the time synchronization packet PKT A relayed on route A is relayed via route B, the residence time can be accurately measured.

FIG. 7G is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG.
In FIG. 7G, it is assumed that communication I/F units 102-a and 102-d are linked down. In this case, the clock synchronization switching section 306-b connects the clock section 103-b to the clock synchronization section 301-a. At this time, the clock section 103-c synchronizes with the clock section 103-b. This makes it possible to accurately measure the retention time of the time synchronization packets PKT A and PKT B relayed via the communication I/F units 102-b and 102-c.

FIG. 7H is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG. 1;
In FIG. 7H, it is assumed that communication I/F units 102-b and 102-c are linked down. In this case, the clock synchronization switching section 306-a connects the clock section 103-a to the clock synchronization section 301-b. At this time, the clock section 103-d synchronizes with the clock section 103-a. This makes it possible to accurately measure the retention time of the time synchronization packets PKT A and PKT B relayed via the communication I/F units 102-a and 102-d.

FIG. 7I is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG.
In FIG. 7I, it is assumed that communication I/F units 102-b and 102-d on path B are linked down. In this case, the clock synchronization switching section 306-a connects the clock section 103-a to the clock synchronization switching section 301-a, and synchronizes the clock section 103-c with the clock section 103-a. As a result, even when the time synchronization packet PKT B relayed on route B is relayed via route A, the residence time can be accurately measured.

FIG. 7J is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG.
In FIG. 7J, it is assumed that communication I/F units 102-a and 102-b are linked down. In this case, since the time synchronization packets PKT A and PKT B cannot be received and the time cannot be synchronized at all, there is no need to switch synchronization.

FIG. 7K is a diagram showing still another example of the synchronization method of the clock unit when a link down occurs in the synchronization system of FIG.
In FIG. 7K, it is assumed that communication I/F units 102-c and 102-d are linked down. In this case, communication I/F units 102-a and 102-b perform time synchronization with time synchronization servers 104-A and 104-B by time synchronization packets PKT A and PKT B, respectively. However, since the communication I/F units 102-c and 102-d cannot relay the time synchronization packets PKT A and PKT B, it is necessary to synchronize the clock units 103-c and 103-d in order to measure the retention time. There is no

FIG. 8 is a block diagram showing the operation of the clock synchronization switching unit of FIG.
In FIG. 8, the clock synchronization switching unit 306-a includes switches W1 to W3 for switching paths. The clock synchronization switching unit 306-b includes switches W4 to W6 for switching paths.

When the switch W1 is turned on, the clock synchronization switching section 306-a uses the clock section 103-a as the reference source clock for synchronization. When the switch W2 is turned on, the clock synchronization section 301-b uses the clock section 103-a as a reference source clock for synchronization. When the switch W3 is turned on, the clock synchronization section 301-a synchronizes the time of the clock section 103-a.

When the switch W4 is turned on, the clock synchronization section 301-b uses the clock section 103-b as a reference source clock for synchronization. When the switch W5 is turned on, the clock synchronization section 301-a uses the clock section 103-b as a reference source clock for synchronization. When the switch W6 is turned on, the clock synchronization section 301-b synchronizes the time of the clock section 103-b.

Further, the clock synchronization unit 301-a always synchronizes the time of the clock unit 103-c, and the clock synchronization unit 301-b always synchronizes the time of the clock unit 103-d.

FIG. 9 is a diagram showing the synchronization source of the switch and the clock unit according to the link status of the clock synchronization switching unit of FIG. TS-A and TS-B indicate time servers 104-A and 104-B, respectively. Clock units a to d indicate clock units 103-a to 103-d, respectively.

In FIG. 9, synchronization with the time synchronization server when communication I/F units 102-a to 102-d are linked down, synchronization source of clock units 103-a to 103-d when switches W1 to W6 are ON/OFF etc.

For example, when communication I/F section 102-a is linked down, synchronization with time server 104-A cannot be performed. At this time, the switches W4 and W5 are turned ON. As a result, the clock unit 103-a does not synchronize with the time server 104-A, the clock unit 103-b synchronizes with the time server 104-B, and the clock units 103-c and 103-d synchronize with the clock unit 103-b. Synchronize.

Also, when the communication I/F section 102-d is linked down, the synchronization with the time server 104-B is stopped and the switches W1, W2 and W6 are turned ON. As a result, the clock section 103-a is synchronized with the time server A, and the clock sections 103-b to 103-d are synchronized with the clock section 103-a.

FIG. 10 is a diagram showing a time synchronization sequence of the synchronized imaging system of FIG. 10, the time server 104-A and the camera adapters 201-a and 201-b of FIG. 2 are shown. As for the camera adapter 201-a, only the TC function is shown, but like the camera adapter 201-b, the camera adapter 201-a also transmits PTP Delay_Req and synchronizes its own time.

In FIG. 10, the time server 104-A transmits a PTP SYNC packet (p101) and stores the transmission time T1 of the PTP SYNC packet.

Next, the camera adapter 201-a receives the PTP SYNC packet from the time server 104-A and relays the PTP SYNC packet as a TC terminal (p102). At this time, the camera adapter 201-a obtains and stores the dwell time dT1a from the transmission time and reception time of the PTP SYNC packet.

Next, the camera adapter 201-b receives the PTP SYNC packet from the camera adapter 201-a and stores the reception time T2b.

Next, the time server 104-A transmits a PTP Follow_Up packet carrying the transmission time T1 of the PTP SYNC packet (p103).

Next, camera adapter 201-a, as a TC terminal, adds the retention time dT1a of the PTP SYNC packet to a specific area of the PTP Follow_Up packet received from time server 104-A. Then, the camera adapter 201-a relays the PTP Follow_Up packet with the dwell time dT1a added to the specific area (p104).

Next, camera adapter 201-b receives the PTP Follow_Up packet sent from camera adapter 201-a. Then, the camera adapter 201-b determines the transmission time T1 of the PTP SYNC packet on the clock of the time server 104-A and the time from when the PTP SYNC packet is transmitted from the time server 104 to when it reaches the camera adapter 201-b. get.

Next, the camera adapter 201-b transmits a PTP Delay_Req packet and stores the transmission time T3b of the PTP Delay_Req packet (p105).

Next, the camera adapter 201-a relays the PTP Delay_Req packet transmitted from the camera adapter 201-b to the time server 104-A (p106). At this time, the camera adapter 201-a obtains and stores the retention time dT3a from the reception time and the transmission time T3b of the PTP Delay_Req packet.

Next, the time server 104-A receives the PTP Delay_Req packet transmitted from the camera adapter 201-a and stores the reception time T4b. Then, the time server 104-A responds to the PTP Delay_Req packet with a PTP Delay_Resp packet carrying the reception time T4b (p107).

Next, the camera adapter 201-a adds a dwell time dT3a to a specific area of the PTP Delay_Resp packet received from the time server 104-A and relays it (p108).

Next, the camera adapter 201-b receives the PTP Delay_Resp packet transmitted from the camera adapter 201-a. Then, the time taken for the PTP Delay_Req packet transmitted by the camera adapter 201-b to reach the time server 104 and the time of the time server 104 when the PTP Delay_Req packet arrived at the time server 104-A are obtained.

Next, the time server 104-A transmits the PTP SYNC packet (p109) and stores the transmission time T5 of the PTP SYNC packet.

Next, the camera adapter 201-a receives the PTP SYNC packet from the time server 104-A and relays the PTP SYNC packet as a TC terminal (p110). At this time, the camera adapter 201-a obtains and stores the dwell time dT5a from the transmission time and reception time of the PTP SYNC packet.

Next, the camera adapter 201-b receives the PTP SYNC packet from the camera adapter 201-a and stores the reception time T6b.

Next, the time server 104-A transmits a PTP Follow_Up packet carrying the transmission time T5 of the PTP SYNC packet (p111).

Next, camera adapter 201-a, as a TC terminal, adds the retention time dT5a of the PTP SYNC packet to a specific area of the PTP Follow_Up packet received from time server 104-A. Then, the camera adapter 201-a relays the PTP Follow_Up packet with the dwell time dT5a added to the specific area (p112).

Next, camera adapter 201-b receives the PTP Follow_Up packet sent from camera adapter 201-a. Then, the camera adapter 201-b determines the transmission time T5 of the PTP SYNC packet on the clock of the time server 104-A and the time from when the PTP SYNC packet is transmitted from the time server 104 to when it reaches the camera adapter 201-b. get.

Average transmission path delay time TDV (not including dwell time) between time server 104-A and camera adapter 201-b is obtained as follows.
TDV = (((T4b-T1)-(T3b-T2b))-(dT1a+dT3a))/2
The time difference TS between the time server 104-A and the camera adapter 201-b is obtained as follows.
TS=T6b-T5-dT5a-TDV

The calculation of the time difference TS is performed after receiving the PTP Follow_Up packet, which gives the value of the transmission time T5 and dwell time dT5a of the PTP SYNC packet. Time correction is performed based on this time difference TS. The retention time (dT1a, dT5a, etc.) required for the relay processing within the camera adapter 201-a is not constant due to transmission/reception buffers and relay processing. Therefore, it is necessary to measure the residence time for accurate time adjustment.

FIG. 11 is a flow chart showing the operation of the camera adapter of FIG.
Each step in FIG. 11 is implemented by the CPU reading and executing the program stored in the camera adapter 201 in FIG. Also, at least part of the flowchart shown in FIG. 11 may be implemented by hardware. When implemented by hardware, for example, by using a predetermined compiler, a dedicated circuit may be automatically generated on an FPGA from a program for implementing each step. Also, a Gate Array circuit may be formed in the same manner as the FPGA and implemented as hardware. Also, it may be realized by an ASIC.
In this case, each block in the flowchart shown in FIG. 11 can be regarded as a hardware block. A plurality of blocks may be collectively configured as one hardware block, or one block may be configured as a plurality of hardware blocks.

In FIG. 11, the camera adapter 201 adjusts the clock units 103-a to 103-d to either time of the time server 104-A or 104-B according to the link state, and at the same time relays the time synchronization packet. . In this flowchart, the time synchronization control units 304-a and 304-b are configured as software tasks.

The camera adapter 201 activates PTP synchronization/relay tasks, which are the time synchronization control units 304-a and 304-b (s100).
Next, if the camera adapter 201 does not receive the time synchronization packet (s101-N), it returns to s101. When the camera adapter 201 receives the time synchronization packet (s101-Y), the abnormality recovery detection units 303-a to 303-a to 102-d detect whether the link states of the four communication I/F units 102-a to 102-d have changed. 303-d is used to check (s102). When the link state changes (s102-Y), the camera adapter 201 refers to FIG. 6 to check the relay destination of the packet (s103), and proceeds to s107. If the link state has not changed (s102-N), the camera adapter 201 proceeds to s107.

If the received time synchronization packet is a PTP SYNC packet (s107-Y), the camera adapter 201 acquires the time when the time synchronization packet was received (reception time stamp) (s112) and relays the packet. In this relay process, the camera adapter 201 transmits the PTP SYNC packets received by the communication I/F units 102-a and 102-b from the communication I/F units 102-c and 102-d (s113). Then, the camera adapter 201 acquires the time when the time synchronization packet was transmitted (transmission time stamp) (s114), calculates the retention time (s115), and returns to s101.

If the received time synchronization packet is not a PTP SYNC packet (s107-N), the camera adapter 201 checks whether it is a PTP Delay_Req packet (s1108). If the received time synchronization packet is a PTP Delay_Req packet (s108-Y), the camera adapter 201 performs the processes of s112 to s115 and returns to s101. However, the direction of relaying in the relaying process of s113 is reversed. That is, in this relay process, the camera adapter 201 transmits the PTP Delay_Req packets received by the communication I/F units 102-c and 102-d from the communication I/F units 102-a and 102-b.

If the received time synchronization packet is not a Delay_Req packet (s108-N), the camera adapter 201 checks whether it is a PTP Delay_Resp packet (s109). If the received time synchronization packet is a PTP Delay_Resp packet (s109-Y), the camera adapter 201 checks whether it is addressed to its own host (s110). If it is addressed to its own host (s110-Y), it calculates the average transmission line delay time TDV (s116) and returns to s101. If the packet is not addressed to its own host (s117-N), the camera adapter 201 adds the retention time calculated when the PTP Delay_Req packet was received and performs relay processing. In this relay processing, the camera adapter 201 transmits the PTP Delay_Resp packets received by the communication I/F units 102-a and 102-b from the communication I/F units 102-c and 102-d (s117).

If it is not a PTP Delay_Resp packet in s109 (s109-N), the camera adapter 201 checks whether the received time synchronization packet is a PTP Follow_Up packet (s111). If it is not a PTP Follow_Up packet (s111-N), the camera adapter 201 returns to s101. If the packet is a PTP Follow_Up packet (s111-Y), the camera adapter 201 adds the retention time calculated when the PTP SYNC packet was received, and relays the PTP Follow_Up packet. In this relay process, the camera adapter 201 transmits the PTP Follow_Up packets received by the communication I/F units 102-a and 102-b from the communication I/F units 102-c and 102-d (s118). Then, the camera adapter 201 performs clock control selection processing as shown in FIG. 8 (s120).

As a result of the clock control selection process, if the time servers 104-A and 104-B are synchronizing with the clock units 103-a and 103-b, the camera adapter 201 receives the time synchronization packet and the clock units 103-a and 103-b. , the process returns to s101 (s121-Y). If the time servers 104-A and 104-B are not synchronized with the clock units 103-a and 103-b (s121-N), the camera adapter 201 sends PTP Delay_Req packets to the time servers 104-A and 104-B. Send (s122). Then, the camera adapter 201 performs processing for synchronizing the times of the clock units 103-a and 103-b with the time servers 104-A and 104-B (s119).

When performing time synchronization using network communication, after observing the time difference TS with the time servers 104-A and 104-B, the progress of the time is adjusted, resulting in gradual fluctuations. Here, when relaying the time synchronization packet, the camera adapter 201 synchronizes the clock section of the communication I/F section that receives the time synchronization packet with the clock section of the I/F that is the relay destination of the time synchronization packet. As a result, the camera adapter 201 can eliminate fluctuations for adjusting the advance of the time, and can more accurately measure the retention time required for relaying the time synchronization packet.

<Second embodiment>
FIG. 12 is a diagram showing synchronization sources of switches and clock units according to link status of the clock synchronization switching unit according to the second embodiment. FIG. 12 is a part of FIG. 9 and shows only the parts different from FIG.
In FIG. 12, for example, when the communication I/F unit 102-a is linked down, not only are the switches W4 and W5 of the clock synchronization switching unit 306-b in FIG. The switch W3 of -a is also turned ON. Thereby, the clock section 103-a can be synchronized with the clock section 103-b.

With this configuration, when the times of the time servers 104-A and 104-B are very close to each other, the time required to restore synchronization with the time servers 104-A and 104-B when the link is restored can be shortened. .

<Third Embodiment>
In the abnormality detection of the abnormality recovery detection unit 303-a of the first embodiment, when the packets transmitted by the time servers 104-A and 104-B are both received by the same communication I / F unit, it is assumed that the link has gone down .

As a result, the link down (communication failure) of the communication I/F unit 102-c of the camera adapter 201-b in FIG. 4 was detected, but the link down of the communication I/F unit 102-a of the camera adapter 201-c was It can be dealt with if it cannot be detected. Such a situation occurs because, for example, a part of the communication I/F unit 102-c of the camera adapter 201-b has failed, but since it is electrically normal, the communication of the camera adapter 201-c is not possible. This is the case where the I/F unit 102-a cannot detect it.

<Fourth Embodiment>
FIG. 13 is a flow chart showing the operation of the clock synchronizer according to the fourth embodiment.
In FIG. 13, when the link down of the communication I/F section occurs, the camera adapter 201 changes the synchronization source of the clock section of the communication I/F section. At this time, if the synchronization destination clock unit does not generate a Genlock signal for synchronizing image capture, the synchronization source is synchronized with the synchronization destination time, and then time synchronization is performed to generate the Genlock signal. , time synchronization is performed by adjusting the frequency of the clock unit.

That is, the camera adapter 201 activates the clock synchronization unit (s300) and confirms whether or not the synchronization destination is generating the Genlock signal (s301). If the synchronization destination does not generate the Genlock signal (s301-N), the time of the synchronization source clock section is set in the synchronization destination clock section (s302), and the synchronization destination clock section is set based on the time of the synchronization source clock section. The frequency of the time on the clock unit is adjusted (s303). If the synchronization destination clock unit generates a Genlock signal (s301-Y), the time frequency adjustment is performed without setting the time of the synchronization source clock unit to the synchronization destination clock unit (s303).

Next, the camera adapter 201 waits, for example, one second (s304) and checks whether the link state has changed (s305). Then, the camera adapter 201 returns to s301 if the link state has changed (s305-Y), and returns to s303 if the link state has not changed (s305-N).

In this configuration, when the time of the clock of the synchronization destination differs greatly from the time of the clock of the synchronization source (for example, at startup), the time of the clock of the synchronization source is set to the clock of the synchronization destination. Since time synchronization is performed later by frequency adjustment, the time until time synchronization can be shortened. Further, when the clock section generates the Genlock signal, the time is adjusted by adjusting the frequency without setting the synchronization target time, so the Genlock signal is not disturbed.

In this embodiment, the Genlock signal is used as an example, but the present invention is not limited to this. It is determined that the clock unit is being used when the clock unit is referenced within a certain period of time, and time synchronization is performed by adjusting the frequency. good too.

<Other embodiments>
The present invention may supply a program that implements one or more functions of the above-described embodiments to a system or device via a network or storage medium. One or more functions of the above-described embodiments can also be realized by a process in which one or more processors in the computer of the system or device read and execute the program. It can also be implemented in a circuit (eg FPGA or ASIC) that implements one or more functions.

102-a to 102-d communication I/F section, 103-a to 103-d clock section, 104-A, 104-B time server, 200 synchronous imaging system, 201 camera adapter, 301-a, 301-b clock Synchronization unit, 303-a to 303-d abnormality recovery detection unit, 304-a, 304-b time synchronization control unit, 305-a to 305-d path switching unit, 306-a, 306-b clock synchronization switching unit

Claims (11)

A first communication interface used for relaying the time synchronization packet on the first path;
A second communication interface used for relaying the time synchronization packet on the second path;
a first timer provided in the first communication interface;
a second timer provided in the second communication interface;
Synchronization means for synchronizing the first clock means and the second clock means based on the link state of at least one of the first path and the second path;
A communication device comprising: The first timekeeping means includes a first time stamp acquisition means for acquiring a time stamp of transmission and reception of the time synchronization packet of the first communication interface, based on the time stamp acquired by the first time stamp acquisition means, Measuring the residence time required for relaying the time synchronization packet by the first communication interface,
The second timekeeping means includes a second time stamp acquisition means for acquiring a time stamp of transmission and reception of the time synchronization packet of the second communication interface, based on the time stamp acquired by the second time stamp acquisition means, 2. The communication device according to claim 1, wherein a retention time required for relaying said time synchronization packet by said second communication interface is measured. The synchronization means, based on the link state of at least one of the first path and the second path, the first timing means of the first communication interface or the second communication interface from which the time synchronization packet is received and said second clocking means of said first communication interface or said second clocking means of said second communication interface for relaying said time synchronization packet. 3. The communication device according to 2. detection means for detecting link states of the first route and the second route;
The relay destination of the time synchronization packet is switched between the first route and the second route based on the link state of at least one of the first route and the second route detected by the detecting means. a route switching means;
The synchronization destination of the time of the synchronizing means is selected between the first time measuring means and the second time measuring means based on the link state of at least one of the first path and the second path detected by the detecting means. a synchronous switching means for switching between
4. The communication device according to any one of claims 1 to 3, comprising: a first time synchronization server that synchronizes with the first clock means provided in the first communication interface;
a second time synchronization server that synchronizes with the second clock means provided in the second communication interface;
5. A communication device according to any one of claims 1 to 4, comprising: When the relay destination of the time synchronization packet is changed from the first path to the second path, the first communication interface stops time synchronization between the first timekeeping unit and the first time synchronization server,
The second communication interface stops time synchronization between the second clock unit and the second time synchronization server when the relay destination of the time synchronization packet is changed from the second path to the first path. 6. The communication device according to claim 5, characterized by: The first communication interface synchronizes the time of the first timekeeping unit with the second timekeeping unit that is time-synchronized with the second time synchronization server when the first communication interface becomes unable to communicate,
The second communication interface is characterized by time-synchronizing the second timekeeping unit with the first timekeeping unit that is time-synchronized with the first time synchronization server when the second communication interface becomes unable to communicate. 7. The communication device according to claim 5 or 6, wherein When the first communication interface receives the time synchronization packet transmitted from the second time synchronization server, it is determined that the second communication interface has become unable to communicate,
8. Any one of claims 5 to 7, characterized in that said second communication interface determines that said first communication interface has become unable to communicate when it receives a time synchronization packet transmitted from said first time synchronization server. 1. The communication device according to claim 1. 9. The communication apparatus according to any one of claims 1 to 8, wherein said synchronizing means gradually synchronizes said first clocking means and said second clocking means. a step of relaying the time synchronization packet via a first communication interface on a first path provided with a first timing means;
a step of relaying the time synchronization packet via a second communication interface on a second path provided with a second timing means;
synchronizing the first timing means and the second timing means based on the link state of at least one of the first path and the second path;
A communication method comprising: A program for operating a computer as the communication device according to any one of claims 1 to 9.

Images