Scheme of splicing screen of traffic police command center and multi-channel monitoring wall

Translate "how many roads are connected" into "how many roads can be seen clearly at the same time": signal classification, decoding conversion, bandwidth accounting, layout plan and full load acceptance, a calculable link

The most expensive mistake of the traffic police command center wall plan is to regard "access to 1000 roads" as "being able to see 1000 roads at the same time".How many channels are registered on the platform, how many channels can be forwarded by the platform, how many channels can be solved by the decoder, and how many channels can be clearly seen on the screen-these four numbers are completely different. Any wrong calculation in any link will be exposed during acceptance: either the decoder is not enough, and Caton has a black screen during full screen switching; Either there are too many windows and each picture is too small to see the license plate clearly.The multi-channel monitoring wall is not "connecting the line", but a capacity chain from the signal source to the human eye, and each layer must be accounted.This article focuses on this main technical line: to clarify the unique signal classification of traffic police, to convert the number of decoding paths and bandwidth into applicable tools, and to put the layout plan and full load acceptance into testable terms.This paper forms a series with "the key points of the design of the LCD splicing screen of the public security command center"-- General methodologies such as on-site survey, reverse pushing of the number of rows and columns on the screen wall, centralized and distributed trade-offs, installation and operation, etc., please refer to the Public Security section; this article only expands the calculation and judgment unique to the "wall link.

Six types of signals unique to 1. traffic police: what link to take and what equipment to match.

Many wall-up schemes treat "all signals" as "video stream decoding". As a result, there are too many decoders, the service system interface is not left, and the bayonet picture cannot be wall-up. The signal sources of the traffic police command center can be divided into two categories and six types according to the wall link.

Special requirements for traffic police service signals

In addition to the general video wall, traffic police have several types of signals that have special requirements for wall-to-wall methods, which are easy to be ignored in the scheme:

• bayonet snap pictures: Usually a large-resolution static image (such as 2 million -9 million pixels), not a video stream. When going to the wall, confirm whether the business system supports the "picture carousel" mode to output to the large screen; Whether the license plate number is readable after a single picture is enlarged on the large screen; Whether the picture refresh is actively pushed or pulled regularly. These decisions are made by the cooperation of the business system interface and the display terminal, not by the decoder.
• Signal timing scheme and phase diagram: Usually in the form of business system pictures (tables or graphics) on the wall for the commander to study and judge. To confirm the "screen refresh frequency" and "whether to support multi-road on-screen comparison", rather than video decoding capabilities.
• Traffic Guidance Information: The large screen shows the monitoring screen of "the current release content of the induction screen,Show Not Equals Control-- The ability to publish induced information belongs to the business platform and authorization interface, and integrators should not be required to promise that "large screen can change induced screen content".
• 122 Police Work Order/Accident Site Chart: mostly for business system interface or GIS annotation screen, press the application screen on the wall. What needs to be confirmed is whether the operating end can be retrieved quickly and whether it can be compared with the video screen on the same screen.

How to convert the number of 2. decoding paths: three variables and one table

The "nominal number of ways" of the decoder is usually the full capacity at a particular encoding format, at a particular resolution. The real needs of the project must be converted according to the "coding format x resolution x frame rate", can not directly take the nominal number of paths as the actual capacity.

The three variables for conversion

Encoding format: The compression efficiency of H.265 is higher than that of H.264, and the code rate is lower under the same picture quality. The number of ways for the same decoder to solve H.265 is usually higher than that of H.264. But the specific difference depends on the decoding chip, different models vary greatly, can not be generalized. The plan shall require the supplier to provide "the number of decoding paths in the actual encoding format of the project".

Resolution: The number of pixels in a 4K frame is 4 times that of 1080P, and the computing power occupied by decoding 4K is also much higher than 1080P. The common capability caliber is "4 channels 4K or 16 channels 1080P decoding at the same time"-this 4:16 ratio (I .e. 1 channel 4K ≈ 4 channels 1080P) is the conversion reference for many decoders, but the specific model specification shall prevail.

Frame rate: Monitor the influence of the common 25fps and 30fps on the decoding pressure, which is usually less than the influence of the resolution, but will become a variable in the full edge state. The scheme should confirm "decoding capability at the target frame rate" instead of just looking at the label "4K supported.

The items with mixed resolution shall be weighted according to the conversion coefficient. For example, a project has 4 roads 4K, 20 roads 1080P and 10 roads 720P that need to be mounted on the wall at the same time, and the conversion requirement is about 4 × 4 20 × 1 10 × 0.5 = 41 road 1080P equivalent. If a single decoder is fully equipped with 16 channels 1080P, at least 3 decoders are required (and confirm the number of output interfaces covering the number of screens).

3., the "display position" is deduced from the number of screens and the number of partitions: the upper limit of decoding requirements is determined by the screen wall.

The number of decoding channels is not determined by the "total number of cameras", but by the "how many windows the screen wall can display at the same time. This window count represents the total number of “display positions.”

Calculation of display position

This calculation connected the "screen wall size" with the "decoding requirements:The more the number of screens and the finer the division, the more the number of decoding channels are required; but the finer the division, the smaller the single-channel picture, and the worse the readability. Therefore, the display position is not the more the better, but to balance between "covering the number of ways" and "single-way readability. Key intersections are divided by 4 (the picture is large and can be seen clearly) and non-key road sections are divided by 9 or 16 (covering more and seeing the whole), which is a common division strategy for traffic police command centers.

Correspondence between partitioning policy and decoding requirements

In this example, the video display position is 16 54=70 (the map area goes through the application link and does not account for decoding). If these positions need to have pictures at the same time, the decoder must be able to solve at least 70 1080P equivalent simultaneously. If a round trip is used (different pictures are displayed in the same position), the decoding requirement can be lower than the number of display positions--However, the round-robin does not increase the number of "simultaneously visible" routes., it just allows limited positions to take turns showing more pictures.

4. bandwidth accounting: The network pressure of concurrent stream fetching is easier to ignore than decoding.

The decoder is right. If the network bandwidth is not enough, there will be problems. Multi-channel video streams are taken from the platform at the same time, and the bandwidth requirement after the code rate is superimposed often exceeds the estimate of the scheme stage.

Estimation of Bit Rate and Bandwidth

Bandwidth accounting can not only look at the "decoder to switch link", but also lookConcurrent forwarding capability on the platform side. The forwarding capability of the video networking platform is limited by authorization and server performance. In the scheme, the supplier shall be required to confirm the forwarding capability of the platform under the target number of concurrent paths, whether there is a concurrent authorization limit, and whether the main stream/sub-stream switching is available. If the forwarding capacity of the platform is insufficient, the decoder cannot get the stream no matter how strong it is.

5. project scale difference: municipal detachment, district brigade, high-speed traffic police, DMV is not the same set of configuration

"Traffic police command center" is a broad term, and the scale and configuration of different levels vary greatly. The same set of scheme templates is applied to all levels, either big horse-drawn cart or small horse-drawn cart.

The difference in scale directly changes the focus of the wall-up scheme: municipal detachments are most afraid of "multi-level cascading and platform concurrent sticking", district brigades are most afraid of "too much waste in decoding and not enough in matching", high-speed traffic police are most afraid of "abnormal display of long-distance transmission and bayonet pictures", and DMV is most afraid of "too complicated scheme and inability to keep up with operation and maintenance". The scheme design should first confirm which category the project belongs to, and then determine the configuration benchmark of decoding capacity, bandwidth and architecture.

6. layout and plan: let the big screen follow the peak inspection and accident disposal.

The large screen of the traffic police command center does not "always display the same content", but switches the layout with the task. The layout and plan design should focus on three typical working modes.

Three working modes of the layout of the draft

• Morning and evening peak patrol mode: Fixed windows at key intersections are often displayed (4 divisions, large picture) and wheel patrol areas (9 divisions, wheel patrol according to jurisdiction). Situation billboards are often displayed. The plan should support "one-click switch to peak mode", and the round-trip grouping should be preset according to the jurisdiction and time period.
• Accident disposal mode: The accident point screen is enlarged (single screen or less divided) and the accident location is marked on the GIS map of the linkage window of the upstream and downstream intersection. The plan should support "one-click access to surrounding videos from map points", but this capability depends on platform and interface support and cannot be written as a self-contained function of the screen.
• Major event guarantee mode: Multi-channel fixed monitoring (key road sections, venue surroundings) situation billboard video conference screen on the same screen. The plan should support the same screen and fast switching of multi-department signals.

Three measured items of plan acceptance

The plan is not "just configured" and should be measured during acceptance:Plan switching time(From operation to picture stability, whether it meets actual combat requirements);Black field/pause for round-robin switching(Whether there is obvious black screen or Caton at the moment of switching, and whether it is acceptable);Response of alarm linkage screen(If the event triggers the corresponding picture on the large screen, whether the delay is within an acceptable range). These three items shall be based on the measured values agreed in the contract, not the nominal values of the supplier.

7. Reliability and Restoration from Disconnection: Fault Scenarios peculiar to Wall Link

For the 7 × 24 traffic police command center, in addition to the redundancy of general equipment (dual power supply and processor hot standby, see the failure recovery section of the public security chapter for details), there are several unique failure scenarios for the upper wall link that need to be designed for recovery schemes.

The acceptance of disconnection recovery should not only look at "whether it can be recovered after disconnection", but also test "how long it will take to recover, whether the picture after recovery is consistent with that before disconnection, and whether an error alarm will be triggered". These indicators should be included in the acceptance outline as a benchmark for the warranty period.

8. full load acceptance: lock every promise with real load

"Demonstration is normal, upper line pull crotch" is the most common way to roll over the wall project. During the demonstration, only a few signals were received and only a few windows were opened. When the acceptance was full, the problem was exposed. The purpose of full load acceptance is to run the real load again in the acceptance stage.

Test items for full load acceptance

The environment for full load acceptance should be consistent with the real environment: the same code stream, the same number of windows, and the same network load. If the light load "dedicated to demonstration" is used for acceptance, it is meaningless to pass the acceptance. The acceptance outline shall be written in the annex to the contract, and each indicator shall correspond to "evidence of delivery of test conditions and test methods" to avoid vague expressions such as "good results.

9. Quotation Boundary: Five Items Most Easily Missed by Traffic Police on Wall

If the quotation of the traffic police on the wall project is only higher than the "screen unit price" or "decoder unit price", the later increase will almost inevitably occur. The following five items are the most easily missed boundaries in the quotation.

• Platform docking and joint debugging costs. The interface between the decoder and the video networking platform, bayonet system and GIS platform may require interface authorization on the platform side, account opening and cooperation from third-party manufacturers. It should be clear who will bear this part of the workload and whether it is included in the quotation.
• Number of output interfaces of the decoder and board configuration. The decoder nominally "supports 16-way decoding", but there may be only 4 or 8 output interfaces. When the number of screens is more than the number of output interfaces, it is necessary to increase the output board or cascade equipment, which is easy to be omitted.
• Network support. Whether the switch, optical fiber/network cable, optical module and cabinet space are included in the quotation. If the network is provided by the owner, the interface boundaries and responsibilities for joint testing must be clearly defined.
• Plan preparation and optimization. Debugging should not only include "signal on, picture on", but also include the plan configuration and optimization of the three working modes. The workload for preparing the plan and the number of revision cycles shall be specified in the quotation.
• Warranty starting and spare parts supply. The warranty starts from "acceptance" or "delivery"; whether the spare parts list (decoder board, power supply, fan) is included; the supply life of spare parts for the panel/module. These directly affect long-term costs.

10. common wall scheme misunderstanding

• Misunderstanding 1: Take "the number of access roads" as "the number of roads on the wall at the same time". The platform is connected to thousands of roads, which does not mean that the large screen can display thousands of roads at the same time. At the same time, the ability to go up the wall is constrained by the number of decoding paths, platform forwarding concurrency, and network bandwidth, and the minimum value of the three is taken.
• Myth 2: Decoders are purchased according to the total number of cameras. Decoding requirements should be configured according to the display position calculated by "screen number x division number", and then converted according to encoding format and resolution, not according to the total number of cameras. Access 800 way does not mean that 800 way decoding is required.
• Myth 3: Thinking that rounds can increase the number of roads monitored at the same time. The round trip is time-sharing switching, and the number of windows that can be seen at the same time remains unchanged. What the rounds expand is "inspection coverage", not "simultaneous monitoring capability".
• Myth 4: Put all signals into the decoder as a video stream. The bayonet picture, the situation dashboard, and the GIS map are application pictures, which follow the output link of the business system and do not occupy decoding resources. Unclear classification will result in multiple decoders or no service interface left.
• Myth 5: Only decoding, not bandwidth and platform concurrency. No matter how strong the decoder is, the platform's forwarding capacity is insufficient or the network bandwidth is insufficient, and the stream cannot be obtained. The three "concurrency" shall be verified separately and the minimum value shall be taken.
• Myth six: acceptance with light load demonstration. During the demonstration, several roads and windows were opened, and the problem was exposed only when the acceptance was full and full. Mixed signal full load test must be done, and the test environment is consistent with the real environment.
• Myth 7: "display" as "control". The large screen can present signal timing and induction information, which does not mean that the signal lamp can be controlled or the content of the induction screen can be modified. The boundary should be clearly written and the acceptance should be measured separately.

XI. FAQ

XII. Conclusion