High-precision time synchronization method for multi-screen tiled display system

Abstract

The invention discloses a high-precision time synchronization method for a multi-screen tiled display system. The high-precision time synchronization method comprises the steps that a master control node broadcasts a global time reference signal regularly; the input node receives video frames from different video sources and prints a local timestamp on each frame of video, and the input node carries out primary correction on the timestamps according to the time reference signal broadcasted by the master control node; the output node receives the video frame subjected to primary correction through the network, and performs secondary correction on the timestamp according to the latest broadcast time reference signal of the master control node; and the output node synchronously triggers the refreshing operation of the plurality of display screens at a preset time point according to the timestamp after secondary correction, so that all the display screens display consistent picture contents at the same moment.

Description

Technical Field

[0001] The present invention relates to the technical field of video processing, and particularly to a high-precision time synchronization method for a multi-screen splicing display system. Background Art

[0002] With the development of video technology and multimedia applications, multi-screen splicing display systems are increasingly widely used in various occasions, such as large conference rooms, exhibitions, monitoring centers, etc. In order to achieve seamless picture splicing effects, high-precision time synchronization must be maintained between individual display screens. However, in existing technologies, due to the influence of factors such as network latency and node clock drift, it is often difficult to achieve ideal synchronization accuracy.

[0003] Therefore, there is an urgent need for a method and technical solution that can effectively solve these problems. Summary of the Invention

[0004] To solve at least one of the foregoing technical problems, the present disclosure proposes, in a first aspect, a high-precision time synchronization method for a multi-screen splicing display system, including: a master node periodically broadcasts a global time reference signal; an input node receives video frames from different video sources and attaches a local timestamp to each video frame, and the input node performs an initial correction on the timestamp according to the time reference signal broadcast by the master node; an output node receives the video frames that have been initially corrected through a network, and performs a secondary correction on the timestamp again according to the latest broadcast time reference signal of the master node; the output node synchronously triggers a refresh operation of multiple display screens at a predetermined time point according to the timestamp after the secondary correction, so that all display screens display consistent picture content at the same moment.

[0005] Preferably, after receiving the broadcast time reference signal from the master node, the input node calculates the deviation between the local clock and the broadcast time, and performs an initial correction on the timestamp of the video frame according to the deviation.

[0006] Preferably, after receiving the latest broadcast time reference signal from the master node, the output node recalculates the deviation between the local clock and the broadcast time, and performs a secondary correction on the timestamp of the video frame according to the deviation.

[0007] Preferably, low-latency switches and routers are connected between the master node and the input node and the output node, and a QoS policy is configured to preferentially transmit the broadcast signal of the master node.

[0008] Preferably, when the local clock of the output node exceeds the timestamp after the secondary correction, the output node immediately refreshes the display screen or discards the frame.

[0009] Preferably, both the input node and the output node include buffers for temporarily storing the received video frames.

[0010] Preferably, the master control node is equipped with a high-precision clock or connected to an external time source.

[0011] In a second aspect, the present disclosure provides a high-precision time synchronization system for a multi-screen splicing display system, including: a broadcast module for the master control node to periodically broadcast a global time reference signal; a primary calibration module for the input node to receive video frames from different video sources, and attach a local timestamp to each video frame, and the input node performs primary calibration on the timestamp according to the time reference signal broadcast by the master control node; a secondary calibration module for the output node to receive the video frames that have been primarily calibrated through the network, and perform secondary calibration on the timestamp again according to the latest time reference signal broadcast by the master control node; a display module for the output node to synchronously trigger the refresh operations of multiple display screens at a predetermined time point according to the secondarily calibrated timestamp, so that all display screens display consistent picture content at the same moment.

[0012] Preferably, the input node and the output node are respectively provided with an anti-shake algorithm module for smoothing network delay fluctuations.

[0013] In a third aspect, the present disclosure provides a computer-readable medium, in which a computer program is stored, and the computer program is loaded and executed by a processing module to implement the steps of any of the above methods.

[0014] Some technical effects of the present disclosure are as follows: A high-precision time synchronization method for a multi-screen splicing display system, including: the master control node periodically broadcasts a global time reference signal; the input node receives video frames from different video sources, and attaches a local timestamp to each video frame, and the input node performs primary calibration on the timestamp according to the time reference signal broadcast by the master control node; the output node receives the video frames that have been primarily calibrated through the network, and performs secondary calibration on the timestamp again according to the latest time reference signal broadcast by the master control node; the output node synchronously triggers the refresh operations of multiple display screens at a predetermined time point according to the secondarily calibrated timestamp, so that all display screens display consistent picture content at the same moment. In the presence of network delays and node clock drifts, it still ensures that data from different video sources can be displayed at the correct time point, and all display screens can synchronously refresh the picture, so that the pictures of each display screen are kept synchronized. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] To better understand the technical solutions of the present disclosure, reference can be made to the following drawings for assisting in the description of the prior art or embodiments. These drawings will selectively show the products or methods involved in the prior art or some embodiments of the present disclosure. The basic information of these drawings is as follows: Figure 1This is a flowchart of an embodiment of a high-precision time synchronization method for a multi-screen splicing display system of the present application. Detailed implementation manners

[0016] The technical means or technical effects involved in the present disclosure will be further described below. Obviously, the provided embodiments (or implementation manners) are only some of the implementation manners covered by the present disclosure, rather than all. Based on the embodiments and the explicit or implicit disclosure in the drawings and text in the present disclosure, all other embodiments that those skilled in the art can obtain without creative efforts will fall within the scope of protection claimed in the present disclosure.

[0017] When videos from different sources are spliced together to form a large picture, if these video sources are not synchronized, viewers may notice that some parts of the picture are ahead of or lag behind other parts, which will seriously affect the viewing experience and destroy visual consistency. In some application scenarios, such as monitoring centers, live broadcasts of large events, or virtual reality environments, video content needs to smoothly transition and continuously play on multiple display screens. If the video sources are not synchronized, the switching between pictures will appear abrupt, affecting the overall effect. For video sources that include audio, if the video and audio are not synchronized, it will greatly affect the user experience. This situation is particularly common in multimedia presentations or home theater setups. In complex audio-visual systems, multiple types of devices may be involved in collaborative work, including but not limited to computers, TV boxes, video recorders, etc. To ensure that all these devices can effectively cooperate and output a unified picture to viewers, synchronization between them is crucial. In professional fields, such as radio and television stations, live broadcasts of concerts, live broadcasts of sports events, etc., precise time control and synchronization are necessary. Any slight delay or unsynchronization may lead to information transmission errors or a decline in the viewing experience. Therefore, synchronization is very critical in the splicing display system because it ensures that multiple display screens can display content simultaneously, forming a seamless large picture. In practical application scenarios, network delays and node clock drifts inevitably exist.

[0018] To address at least one of the above problems, as Figure 1 , the present application discloses a high-precision time synchronization method for a multi-screen splicing display system, including: S10: The master node periodically broadcasts a global time reference signal; S20: The input node receives video frames from different video sources and attaches a local timestamp to each frame of video. The input node performs an initial correction on the timestamp according to the time reference signal broadcast by the master node; S30: The output node receives the initially corrected video frames through the network and performs a secondary correction on the timestamp again according to the latest time reference signal broadcast by the master node; S40: The output node synchronously triggers the refresh operations of multiple display screens at a predetermined time point according to the secondarily corrected timestamps, so that all the display screens display consistent picture content at the same moment.

[0019] By applying the method of the present disclosure, even in the presence of network latency and node clock drift, it is still ensured that data from different video sources can be displayed at the correct time point, and all the display screens can synchronously refresh the pictures, so that the pictures of each display screen are kept in synchronization.

[0020] In one embodiment, the method proposed by the present disclosure is applied to the technical field of distributed systems and is used for high-precision time synchronization of a multi-screen splicing display system. The distributed system includes a master node, multiple input nodes, a switch, multiple output nodes, and multiple splicing screens. The master node, input nodes, and output nodes perform streaming data exchange through the switch. The input and output nodes and the master node are connected in an IP manner. The output node transmits data through an HDMI cable via a network switch. This method not only improves the flexibility of the system but also makes the system easier to expand and maintain. When new input or output devices need to be added, only the corresponding nodes need to be added to the network, without making major adjustments to the existing architecture.

[0021] S10: The master node periodically broadcasts a global time reference signal; In one embodiment, the global time reference signal refers to a time synchronization signal generated and periodically broadcast by the master node, which provides a common time reference point for all the input nodes and output nodes in the system. This time reference signal contains the current absolute time information, such as UTC time, and sometimes also contains other auxiliary information, such as the timestamp format, synchronization period, etc. The master node is equipped with a high-precision clock, such as an atomic clock or a GPS module, which ensures the high accuracy of the time reference of the master node. The master node periodically broadcasts the global time reference signal to all other input nodes and output nodes in the system, so that these input nodes and output nodes can operate based on the same time standard. The master node is also responsible for monitoring the overall time synchronization status of the system and adjusting the broadcast frequency or other parameters as needed.

[0022] The master node first obtains the current absolute time information from its internal high-precision clock or an external time source, such as an NTP server, a GPS module, etc. For example, assume the current time is 2025-04-09 14:05:00.000. The master node encapsulates the obtained time information into a time reference signal in a specific format. This signal usually includes the following: the current absolute time, such as 2025-04-09 14:05:00.000; the timestamp format, such as using UTC time or local time; the broadcast period, that is, how often to broadcast the time reference signal; and other auxiliary information, such as version number, checksum, etc. The master node regularly sends the generated time reference signal to all other input nodes and output nodes in the system according to the set broadcast period. The broadcast can be implemented through a local area network (LAN), and common protocols include the UDP (User Datagram Protocol) or TCP (Transmission Control Protocol). When the broadcast signal is transmitted in the network, it will pass through network devices such as switches. This disclosure uses low-latency switches and other network devices to reduce the processing time of data packets in the network. And a QoS (Quality of Service) policy is configured to preferentially transmit the time reference signal to avoid it being blocked by other traffic. This ensures that the time reference signal can reach all nodes in a timely manner and avoids delays in the transmission of the reference signal.

[0023] The system of this disclosure has relatively high requirements for time synchronization, such as millisecond level or higher, and requires a short broadcast period, such as once per second or more frequently. This can ensure that the time deviation of each node always remains within a small range. If the synchronization accuracy requirement is lower in other usage scenarios, a longer broadcast period can be selected, such as once per minute or longer, to reduce the network load and the workload of the master node. The master node can also dynamically adjust the broadcast period according to the running state of the system. For example, when it detects that the time deviation in the system is large, it can shorten the broadcast period to accelerate the synchronization speed; when the system tends to be stable, it can appropriately extend the broadcast period to reduce the network load. All of these are within the scope of protection of this application.

[0024] S20: The input node receives video frames from different video sources, attaches a local timestamp to each video frame, and the input node makes an initial correction to the timestamp according to the time reference signal broadcast by the master node; Multiple input nodes receive video frames from different video sources respectively. Attaching a local timestamp to each frame of data means recording a timestamp for each frame of data, indicating the time point when the frame arrives at the input node. The initial calibration adjusts the local timestamp to a timestamp based on the global time reference according to the time reference signal broadcast by the master node. Multiple input nodes can receive video frame data streams from various video sources respectively, such as: cameras or camcorders, real-time video data collected. Video encoders, which convert analog signals into digital signals. File players, pre-recorded video files. Network streaming media servers, video streams transmitted over the network. Video frames are usually transmitted in a specific format, and common formats include: H.264 / H.265, efficient video compression formats; RTP / RTSP, real-time transport protocols, commonly used in streaming media applications; JPEG / MJPEG, image formats for frame-by-frame transmission. Regardless of the specific format of the data stream, the input node needs to be able to parse these frames and extract valid video data. The input node receives video frames through a network interface or other physical interfaces (such as HDMI, SDI). The input node is equipped with a buffer, and the buffer is used to temporarily store the received video frames to prevent frame loss caused by bursty data streams.

[0025] The input node reads the current time from its local clock and attaches a local timestamp to each frame of data. For example, assume the local time of the input node is 14:05:00.100. This time is used as the initial timestamp of the video frame. Then, the initial timestamp of video frame A is recorded as 14:05:00.100. The original timestamp of all video frames is 2025-04-09 14:05:00.100, and the timestamp is attached at the moment when the video frame arrives at the input node, otherwise it may cause time deviation.

[0026] After the input node receives the time reference signal from the master node, it compares it with its own local time and calculates the time deviation. For example, if the local time of the input node is 2025-04-09 14:00:00.100 and the broadcast time is 2025-04-09 14:00:00.090, then the time deviation is +0.010 seconds. The input node corrects the timestamp of each video frame according to the calculated time deviation. For example, assume the original timestamp of a video frame is 2025-04-09 14:00:00.100, and the timestamp after the initial calibration should be 2025-04-09 14:00:00.090.

[0027] S30: The output node receives the video frames that have been initially calibrated through the network and performs a secondary calibration on the timestamp according to the latest time reference signal broadcast by the master node again; The output node receives the initially corrected video frames from the input node via a network or a switch. After completing the initial timestamp correction of the video frames, the input node transmits the video frames with the corrected timestamps to the output node via the network. The video frames are usually encapsulated into specific packet formats, such as UDP or TCP, and transmitted via a local area network (LAN).

[0028] During the transmission of the video frames from the input node to the output node, the queuing waiting time, transmission time in network devices such as switches and routers, and the variation of network latency cause inconsistent arrival times of the video frames. Some packets may be lost in the network, resulting in missing video frames. All of these will affect the transmission time of the video frames from the input node to the output node.

[0029] After receiving the video frames, the output node extracts the timestamp information therein and records the actual arrival time of the frames. For example, assume that the corrected timestamp of video frame A is 12:00:00.090, and it arrives at the output node at 12:00:00.115.

[0030] The master node broadcasts a global time reference signal regularly to provide the latest and consistent time reference point for all nodes in the system. For example, assume that the master node broadcasts a time reference signal at 12:00:00.120.

[0031] When the output node receives the time reference signal from the master node, it compares the time reference signal with its local time to calculate the current time deviation. The output node records its local time when it receives the time reference signal. Assume that the local time of the output node is 12:00:00.115. Compare the local time with the broadcast time to calculate the time deviation. The time deviation is: -0.005 seconds. The output node performs a secondary correction on the timestamps of the video frames according to the calculated time deviation to make them based on the latest global time reference.

[0032] After receiving the initially corrected video frames, the output node performs a secondary correction on the timestamps again according to the latest broadcast time reference signal. For example, assume that the timestamp of the video frame received by the output node is 2025-04-09 14:00:00.090, and when the output node receives this frame, its local time is 2025-04-09 14:00:00.115, and the latest broadcast time is 2025-04-09 14:00:00.120, then the time deviation is -0.005 seconds, and the corrected timestamp should be 2025-04-09 14:00:00.095.

[0033] There may be a certain drift or deviation in the local clock of the output node, and it cannot be completely consistent with the time reference of the master node. Through secondary calibration, this deviation can be eliminated, so that the timestamps of all nodes are based on the same standard. Network latency and jitter may cause the actual arrival time of the video frame to be later than its target display time. The actual arrival time is the current time of the local clock of the output node when the video frame arrives at the output node. The target display time is the timestamp time after secondary calibration. Through secondary calibration, the timestamp of the video frame can be adjusted according to the latest broadcast time reference signal, so as to compensate for these delays. If only relying on the initial calibration of the input node, the output node may cause time misalignment problems due to its own clock deviation. Through secondary calibration, the accuracy of time synchronization can be further improved to ensure that the video frame can be displayed at the correct time point.

[0034] S40: The output node synchronously triggers the refresh operations of multiple display screens at a predetermined time point according to the timestamp after secondary calibration, so that all display screens display consistent picture content at the same moment.

[0035] Each video frame carries a timestamp, indicating at which time point the frame should be displayed. After the initial calibration of the input node and the secondary calibration of the output node, the timestamps of all video frames have been adjusted based on the global time reference and have high consistency. All output nodes will extract the timestamp after its secondary calibration from the video frame as the target display time. Suppose the timestamp after secondary calibration of video frame A is 12:00:00.095, then the target display time is 12:00:00.095. The output node will read its local time in real time and compare it with the target display time to calculate the deviation between the two.

[0036] Suppose: The target display time of video frame A is 12:00:00.095. The current time is 12:00:00.090. The time deviation is: +0.005 seconds, which means that video frame A still needs to wait for 0.005 seconds to be displayed. Each display screen has a fixed refresh cycle. For example, 60Hz corresponds to 60 refreshes per second, that is, the interval between each refresh is about 16.67ms. The output node needs to trigger the refresh operation within the next refresh cycle of the display screen to ensure that the picture content can be updated in time.

[0037] If the current time is earlier than the target display time, the output node will wait for a period of time until the target display time is reached before triggering the refresh. If the current time has exceeded the target display time, the output node can choose to immediately refresh the display screen or discard the frame and wait for the next frame.

[0038] For example, the refresh period of the display screen is 16.67 ms. The current time is 12:00:00.090, and the target display time is 12:00:00.095. The time deviation is +0.005 seconds. The output node will wait for 0.005 seconds and then trigger the refresh operation to ensure that video frame A is displayed at the target display time of 12:00:00.095. To achieve a seamless picture splicing effect, all different output nodes must trigger the refresh operation at the same moment.

[0039] The output node triggers the refresh operation of the display screen at a predetermined time point according to the corrected timestamp, ensuring that all display screens display the same content at the same moment.

[0040] In one embodiment, a multi-screen splicing display system includes a master node, two input nodes, and two output nodes. The master node obtains the current time as 2025-04-09 14:05:00.000 from its internal high-precision clock and generates a time reference signal. The master node broadcasts the time reference signal to all input nodes and output nodes regularly at a frequency of once per second.

[0041] Input node 1 receives a frame of video A at 2025-04-09 14:05:00.100 and stamps it with the timestamp 2025-04-09 14:05:00.100. Input node 1 receives the time reference signal from the master node (assumed to be 2025-04-09 14:00:00.090) and calculates the time deviation as +0.010 seconds. The corrected timestamp is 2025-04-09 14:00:00.090.

[0042] The corrected video frame A is transmitted to output node 1 through the network and arrives at output node 1 at 2025-04-09 14:05:00.110. Output node 1 receives the new broadcast signal from the master node (assumed to be 2025-04-09 14:00:00.120) and calculates the time deviation as -0.005 seconds based on its local time (assumed to be 2025-04-09 14:00:00.115). The corrected timestamp is 2025-04-09 14:00:00.095. Output node 1 triggers the refresh operation of the display screen at 2025-04-09 14:00:00.095 according to the corrected timestamp.

[0043] The present disclosure proposes a high-precision time synchronization system for a multi-screen splicing display system in a second aspect, including: a broadcast module for the master node to periodically broadcast a global time reference signal; an input node for receiving video frames from different video sources and attaching local timestamps to each frame of video, and the input node performs an initial correction on the timestamps according to the time reference signal broadcast by the master node; a secondary correction module for the output node to receive the initially corrected video frames through the network and perform a secondary correction on the timestamps again according to the latest broadcast time reference signal of the master node; a display module for the output node to synchronously trigger the refresh operations of multiple display screens at a predetermined time point according to the timestamps after secondary correction, so that all display screens display consistent picture content at the same moment.

[0044] In one embodiment, both the input node and the output node are provided with an anti-shake algorithm module for smoothing the influence brought by network delay fluctuations and improving the time synchronization accuracy. The anti-shake algorithm may include, through multiple samplings, each node will collect the time information in the broadcast signal multiple times and calculate the average value or weighted average value of the multiple samplings. It also includes outlier rejection. If the time deviation of a certain sampling significantly deviates from other sampling values, it is regarded as an outlier and rejected.

[0045] Specifically, the steps of the anti-shake algorithm include: 1. Initialization and parameter setting. Set a reasonable delay fluctuation threshold to distinguish normal network delay changes from abnormal jitters. Set filter parameters: Select a suitable filter type, such as a low-pass filter, Kalman filter, etc., which is not restricted, and adjust its parameters according to system requirements. 2. Data acquisition. Receive data packets. The input node and the output node respectively receive video frame data packets and record the arrival time of each data packet. And assign a local timestamp to each received data packet for subsequent calculation of network delay. 3. Delay detection and quantization. Calculate the delay: For each received data packet, calculate the difference between its actual arrival time and the expected arrival time to obtain the current network delay. Statistically analyze the delay fluctuation: By comparing the delay values of multiple consecutive data packets, statistically analyze the change trend and fluctuation range of the delay. 4. Apply a filter for smoothing. For example, use a low-pass filter and set the cut-off frequency of the filter. Filter the delay value of each data packet to remove high-frequency jitter components. Output the smoothed delay value after filtering. It is also possible to use a Kalman filter, initialize the state vector and covariance matrix. According to the currently observed delay value, predict the next state. Update the state estimate value, combine the predicted value and the actual observed value to obtain a more accurate delay estimate. Use the updated state estimate value for subsequent timestamp correction or refresh operations. 5. Perform timestamp correction based on the smoothed delay value. Input node: Re-evaluate the timestamp correction value of the video frame according to the smoothed delay value. Adjust the timestamp to make it closer to the global time reference. Output node: Optimize the trigger timing of the refresh operation according to the smoothed delay value. Ensure that the video frame can be displayed at the best time point to avoid problems such as frame tearing or other visual issues. 6. Feedback and adjustment. Monitor system performance: Continuously monitor the synchronization accuracy and display effect of the system and collect relevant statistical data. Dynamically adjust parameters: According to the real-time monitoring results, dynamically adjust filter parameters or other control parameters to further improve the stability and anti-interference ability of the system.

[0046] In a third aspect, the present disclosure proposes a computer-readable medium in which a computer program is stored. The computer program is loaded and executed by a processing module to implement the steps of the acquisition method described above. Those skilled in the art can understand that all or part of the steps in the embodiments can be implemented by instructing relevant hardware through a computer program. This program can be stored in a computer-readable medium, and the readable medium can include various media that can store program codes, such as a flash drive, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk, or an optical disc.

[0047] Within the scope of the knowledge and capabilities of those skilled in the art, various embodiments or technical features mentioned herein can be combined with each other as alternative embodiments without conflict. These limited alternative embodiments formed by combining a finite number of technical features, which are not listed one by one, still fall within the technical scope disclosed in the present disclosure and can also be understood or inferred by those skilled in the art in combination with the accompanying drawings and the above text.

[0048] In addition, the descriptions of most embodiments are developed based on different focuses. For a further understanding of the unelaborated parts, reasonable inferences can be made by referring to the relevant content of the prior art, other relevant descriptions herein, or the inventive concept.

[0049] It is emphasized again that the embodiments listed above are relatively typical and preferred embodiments of the present disclosure, which are only used to illustrate and explain the technical solutions of the present disclosure in detail for the convenience of readers' understanding, and are not intended to limit the scope or application of the present disclosure claimed. Any technical solutions obtained by making any modifications, equivalent replacements, improvements, etc. within the spirit and principles of the present disclosure should be covered within the scope of the present disclosure claimed.

Claims

1. A high-precision time synchronization method for a multi-screen splicing display system, characterized in that: include: The master node broadcasts the global time reference signal periodically; The input node receives video frames from different video sources and adds a local timestamp to each frame. The input node performs an initial correction on the timestamp based on the time reference signal broadcast by the master control node. The output node receives the video frame that has been initially corrected through the network, and performs a secondary correction on the timestamp based on the latest broadcast time reference signal from the master control node; The output node synchronously triggers the refresh operation of multiple display screens at a predetermined time point according to the timestamp after the secondary correction, so that all display screens display the same picture content at the same time.

2. The method according to claim 1, characterized in that After receiving the broadcast time reference signal from the master control node, the input node calculates the deviation between the local clock and the broadcast time, and performs an initial correction on the timestamp of the video frame based on the deviation.

3. The method according to claim 1, characterized in that After receiving the latest broadcast time reference signal from the master control node, the output node recalculates the deviation between the local clock and the broadcast time, and performs secondary correction on the timestamp of the video frame according to the deviation.

4. The method according to claim 1, characterized in that The master node is connected to the input node and the output node with low-latency switches and routers, and a QoS policy is configured to give priority to the transmission of the master node's broadcast signal.

5. The method according to claim 1, characterized in that: When the local clock of an output node exceeds the timestamp after the quadratic correction, the output node immediately refreshes the display or discards the frame.

6. The method according to claim 1, characterized in that Both the input node and the output node include a buffer for temporarily storing received video frames.

7. The method according to claim 1, characterized in that The master node is equipped with a high-precision clock or connected to an external time source.

8. A high-precision time synchronization system for a multi-screen splicing display system, characterized in that: include: Broadcast module, used for the master node to periodically broadcast the global time reference signal; The initial correction module is used for the input node to receive video frames from different video sources and add a local timestamp to each frame of video. The input node performs initial correction on the timestamp according to the time reference signal broadcast by the master control node; A secondary correction module is used for the output node to receive the video frame that has been initially corrected through the network, and to perform secondary correction on the timestamp again according to the latest broadcast time reference signal of the master control node; The display module is used for the output node to synchronously trigger the refresh operation of multiple display screens at a predetermined time point according to the timestamp after the secondary correction, so that all display screens display the same picture content at the same time.

9. The system according to claim 8, characterized in that The input node and the output node are respectively provided with an anti-shake algorithm module to smooth the fluctuation of network delay.

10. A computer readable medium, characterized in that: The computer readable medium stores a computer program, which is loaded and executed by the processing module to implement the steps of any one of the methods of claims 1 to 7.

Images