With the construction of Digital China, especially the localization of video surveillance equipment and the rapid decrease in equipment research and manufacturing costs, video surveillance equipment is rapidly becoming popular and widely used in various parks, such as scenic spots, industrial parks, communities, etc. Due to safety and cost reasons, these monitoring devices are distributed in their respective independent intelligent private networks and are in a closed operating state. The main way to access real-time monitoring is through the monitoring command center, which is completed through on-site video monitoring platforms and large screens. In recent years, the rapid popularization of mobile applications on mobile phones has given rise to the demand for remote real-time viewing of video surveillance, such as real-time personnel status at the entrance of scenic spots, on-site videos for emergency management, real-time status of production lines, and the operation of key laboratories. At present, there are many home level Network monitoring camera equipment and systems, which can meet the basic needs of remote monitoring in simple scenes. However, due to cost constraints and camera performance, it is not possible to achieve widespread adoption in the park.
Deploying a cloud based system for video monitoring equipment in the park can meet the requirements of remote monitoring. Real time video streams from the front-end can be obtained through locally deployed NVR devices in the park, and the process of video streaming to the cloud can be completed through transcoding, encryption, and cloud publishing. After downloading, decoding, playing, and other processes on the application end, remote real-time viewing of video monitoring in the park can be completed.
1. Key modules of the system
The key modules of the system work closely together to achieve the goal of cloud and remote monitoring of video surveillance in the park, focusing on the safe and reliable processing, processing, and transmission of video stream data. The following will distribute and introduce the content of each module.
Capture of surveillance video streams
The video monitoring equipment in the park is gathered in the monitoring command center, and real-time video monitoring streams can be obtained through NVR. Surveillance cameras all support the output of multiple bitstreams, including main and sub bitstreams. The encoding formats of the bitstreams include H.264, H.265, etc. The video streams can be obtained through the RTSP protocol. The system first obtains the list details of all video monitoring devices, and then on the configuration page, selects the devices that need to go to the cloud among many cameras, and configures their code stream address, encoding format, account number, password and other information.
In view of the large amount of computing resources and network bandwidth required for simultaneous capture of video streams, the video stream capture part of the system is usually in the "standby" state, and the video stream capture module will only be started when there is a front-end System request request. According to the preset stream configuration information, the front-end real-time stream will be transmitted to the video transcoding platform through NVR. The system platform supports simultaneous capture of multiple video stream data.
Video transcoding
This system needs to support remote viewing on various playback terminals, including PC web pages, WeChat mini programs, and mobile apps. A more general protocol needs to be selected as the transmission standard, so a more general Webrtc format stream has been chosen. To convert the original video stream to Webrtc format according to RTSP standards, it requires two conversions. In the design part of video transcoding, it is completed by two nodes. Firstly, under the control of the self-developed task scheduling component, the FFMEPG component transfers the video stream acquisition module to various RTSP format video streams according to rules, and converts them into corresponding RTMP format streams. Then, the RTMP format streams are converted into corresponding Webrtc format streams through the SRS component according to the scheduling task rules.
Video decoding and playback
The video data has been encrypted, and if it is directly thrown to the player, it cannot be played, and the decryption operation needs to be performed symmetrically. The player obtains encrypted video stream data from the corresponding buffer and hands it over to the decryption module. The decryption module first severely interferes with the correct and consistent release of fields according to the decryption algorithm described in section 1.2 of this article, and then decrypts the remaining data in groups of 4 bytes using the decryption key 0x424E4C49. The decrypted data is directly sent to the player for playback. Thus completing the entire decryption and playback process.
2. System operation effect
After the system is launched and deployed, the transcoding, encryption, decryption, and playback modules seamlessly cooperate and run smoothly. Each video only requires a network bandwidth of 128 kbit/s, and a bandwidth of 20M can support the simultaneous remote retrieval of 200 local monitoring videos in the park. Figure 3 shows the running effect of remote retrieval on the mini program end. The video takes 1.5 seconds from request to playback, runs smoothly, and the screen is clear.
3 Conclusion
The cloud based system for campus video monitoring based on encryption algorithms provides a technical means for remote real-time access of local video monitoring in smart parks. While ensuring timely and efficient remote video monitoring, it also adds a set of barriers to data security. With good commercial value and extensive promotion significance, this system, combined with other quantifiable indicator warning systems, can enable management to effectively grasp the status of generated operations, handle various emergency events in the first time, and effectively promote the digital transformation and upgrading of the park.