In the design and practical application of digital video surveillance systems, it is often necessary to plan the network bandwidth and storage bandwidth of the digital video system. This is very important for the application of the entire system. How to plan the network involves the actual configuration, end use, storage, and editing of the system. Decoding devices, workstations and other distribution factors are the most important in digital video surveillance, but they are more difficult to plan and calculate. First of all need to clear the following concepts:
First, frame rate, stream and resolution
The frame rate concept is a frame of a static picture, continuous frames form an animation, such as television images. We usually say the number of frames, in short, is the number of pictures transmitted in one second, it can also be understood as the graphics processor can refresh several times per second, usually expressed in fps (Frames Per Second). Each frame is a still image. Displaying frames in rapid succession creates an illusion of motion. A high frame rate gives smoother, more realistic animations. The more frames per second (fps), the smoother the displayed action will be.
ï¬ Stream The concept stream (Data Rate) refers to the data traffic used by a video file in a unit of time. It is also called the bit rate. It is the most important part of the picture quality control in video coding. At the same resolution, the larger the code stream of the video file, the smaller the compression ratio and the higher the picture quality.
Resolution Concept Video resolution refers to the size or size of the image formed by the video imaging product. The common video resolution is 352 × 288, 176 × 144, 640 × 480, 1024 × 768. In the two sets of images, the former is the length of the picture, and the latter is the width of the picture. The multiplication of the two results in the pixel of the picture, and the aspect ratio is generally 4:3.
Second, in the digital video system, the main equipment in the bandwidth system is the encoder, network video recorder NVR, and workstations. According to the system, there may be a central server, a management server, a storage archive server, a streaming media server, etc. This article shows a typical architecture. In the system, the network runs mainly upstream video streams, generally UDP, and downlink control commands, which are generally TCP. The main stream comes from the video storage stream from the encoder to the NVR, and the workstation monitors the video stream in real time.
ï¬ Real-time video streaming For real-time video streaming, when the workstation invokes a channel video, the workstation calls different stream types and different code streams. The main consideration is resolution. The average size of the stream at 4CIF resolution can be considered as 1.5M, and encoding The total bandwidth between the device and the NVR depends mainly on the size of each channel stream and the total number of channels. The bandwidth between the workstation and the NVR is mainly determined by the number of video resources called by the workstation. (In this discussion, it is assumed that the encoder does not have multicast capabilities, otherwise there is another thing).
ï¬ Forwarding video stream A function of the NVR is to forward video. The first thing to consider is the output bandwidth of the server. The server uses the network card to output data. In the 100Base-TX switched Ethernet environment, the output bandwidth of a single network card is 1000M. The bandwidth of a video stream is 1.5M, so a 1000M network is sufficient. Therefore, the network card is not a bottleneck. The NVR server's forwarding mechanism, memory, CPU, and video consumption of the overall resources will have an impact on the number of concurrent users.
ï¬ Storage video stream Another function of the NVR is to store video and receive playback commands. This function is mainly used in NVRs and disk arrays. The first thing to consider is the output bandwidth of the server. The other issue to be considered is the output bandwidth and read/write speed of the disk. The main function of the video server is to transmit real-time and playback video data to clients on the network at a stable rate. When there are a large number of concurrent users in the system, the output capacity of the disk is the main bottleneck of the system. A typical disk array can provide 40MB of I/O, a standard MPEG4 video stream bandwidth is 1.5Mbps, so I/O is not a bottleneck, but the NVR server's storage mechanism, memory, CPU, real-time viewing playback, etc., for the overall resources Consumption will affect the number of storage channels.
Third, network video recorder NVR design
ï¬ System memory requirement design
The memory requirements of the NVR system are closely related to the memory occupied by the unit data flow, the size of the data flow rate, the maximum number of streams, and the memory occupied by the streaming media software system itself. In general, the NVR software itself needs to occupy 128MB of available memory. Each 1kbps data stream occupies 12K of memory. The system can use the formula when the maximum number of streams is used: 128M+12×data flow rate×maximum stream number.
ï¬ Number of concurrent users and network bandwidth control technology The network bandwidth is the bottleneck resource in the network. The maximum number of concurrent users that the network can simultaneously support is closely related to the network bandwidth and the bandwidth of the video stream. The network bandwidth can be calculated according to the formula: Network bandwidth = data rate per kbps × maximum number of streams.
ï¬ Server storage space and transmission capacity All video resources of the online video are stored in the video server NVR. The size of the storage space required by the system needs to be determined according to the actual situation of the user. The storage space = number of channels CH × code stream × date of storage. Note: Disk space calculations are rough and cannot be very accurate because the scene of each camera is constantly changing and the code stream is dynamic, so there may be some deviation.
First, frame rate, stream and resolution
The frame rate concept is a frame of a static picture, continuous frames form an animation, such as television images. We usually say the number of frames, in short, is the number of pictures transmitted in one second, it can also be understood as the graphics processor can refresh several times per second, usually expressed in fps (Frames Per Second). Each frame is a still image. Displaying frames in rapid succession creates an illusion of motion. A high frame rate gives smoother, more realistic animations. The more frames per second (fps), the smoother the displayed action will be.
ï¬ Stream The concept stream (Data Rate) refers to the data traffic used by a video file in a unit of time. It is also called the bit rate. It is the most important part of the picture quality control in video coding. At the same resolution, the larger the code stream of the video file, the smaller the compression ratio and the higher the picture quality.
Resolution Concept Video resolution refers to the size or size of the image formed by the video imaging product. The common video resolution is 352 × 288, 176 × 144, 640 × 480, 1024 × 768. In the two sets of images, the former is the length of the picture, and the latter is the width of the picture. The multiplication of the two results in the pixel of the picture, and the aspect ratio is generally 4:3.
Second, in the digital video system, the main equipment in the bandwidth system is the encoder, network video recorder NVR, and workstations. According to the system, there may be a central server, a management server, a storage archive server, a streaming media server, etc. This article shows a typical architecture. In the system, the network runs mainly upstream video streams, generally UDP, and downlink control commands, which are generally TCP. The main stream comes from the video storage stream from the encoder to the NVR, and the workstation monitors the video stream in real time.
ï¬ Real-time video streaming For real-time video streaming, when the workstation invokes a channel video, the workstation calls different stream types and different code streams. The main consideration is resolution. The average size of the stream at 4CIF resolution can be considered as 1.5M, and encoding The total bandwidth between the device and the NVR depends mainly on the size of each channel stream and the total number of channels. The bandwidth between the workstation and the NVR is mainly determined by the number of video resources called by the workstation. (In this discussion, it is assumed that the encoder does not have multicast capabilities, otherwise there is another thing).
ï¬ Forwarding video stream A function of the NVR is to forward video. The first thing to consider is the output bandwidth of the server. The server uses the network card to output data. In the 100Base-TX switched Ethernet environment, the output bandwidth of a single network card is 1000M. The bandwidth of a video stream is 1.5M, so a 1000M network is sufficient. Therefore, the network card is not a bottleneck. The NVR server's forwarding mechanism, memory, CPU, and video consumption of the overall resources will have an impact on the number of concurrent users.
ï¬ Storage video stream Another function of the NVR is to store video and receive playback commands. This function is mainly used in NVRs and disk arrays. The first thing to consider is the output bandwidth of the server. The other issue to be considered is the output bandwidth and read/write speed of the disk. The main function of the video server is to transmit real-time and playback video data to clients on the network at a stable rate. When there are a large number of concurrent users in the system, the output capacity of the disk is the main bottleneck of the system. A typical disk array can provide 40MB of I/O, a standard MPEG4 video stream bandwidth is 1.5Mbps, so I/O is not a bottleneck, but the NVR server's storage mechanism, memory, CPU, real-time viewing playback, etc., for the overall resources Consumption will affect the number of storage channels.
Third, network video recorder NVR design
ï¬ System memory requirement design
The memory requirements of the NVR system are closely related to the memory occupied by the unit data flow, the size of the data flow rate, the maximum number of streams, and the memory occupied by the streaming media software system itself. In general, the NVR software itself needs to occupy 128MB of available memory. Each 1kbps data stream occupies 12K of memory. The system can use the formula when the maximum number of streams is used: 128M+12×data flow rate×maximum stream number.
ï¬ Number of concurrent users and network bandwidth control technology The network bandwidth is the bottleneck resource in the network. The maximum number of concurrent users that the network can simultaneously support is closely related to the network bandwidth and the bandwidth of the video stream. The network bandwidth can be calculated according to the formula: Network bandwidth = data rate per kbps × maximum number of streams.
ï¬ Server storage space and transmission capacity All video resources of the online video are stored in the video server NVR. The size of the storage space required by the system needs to be determined according to the actual situation of the user. The storage space = number of channels CH × code stream × date of storage. Note: Disk space calculations are rough and cannot be very accurate because the scene of each camera is constantly changing and the code stream is dynamic, so there may be some deviation.