A video surveillance device based on edge computing
By introducing edge computing video surveillance devices into the video surveillance system, using hardware-level video codec chips for frame synchronization and bitrate adjustment, and equipped with redundant storage modules, the problems of stuttering and packet loss in video transmission are solved, achieving stable data input and real-time and accurate monitoring, thus improving the reliability of the system.
Patent Information
- Application Number
- CN202522497645.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-25
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2035-11-25
AI Technical Summary
Existing video surveillance systems are susceptible to electromagnetic interference and link problems in park security scenarios, resulting in video transmission lag and packet loss. Edge AI computing boxes lack stable data input, leading to recognition interruptions, numerous monitoring blind spots, and a lack of hardware-level preprocessing and redundancy mechanisms, which fails to meet security traceability requirements.
The video surveillance device, which adopts edge computing, includes multiple campus cameras, edge-side video data preprocessing relay equipment, edge AI computing box and cloud server. It uses hardware-level video codec chip for frame synchronization calibration and adaptive bitrate adjustment, and is equipped with dual-channel redundant storage module to ensure the stability and integrity of data transmission.
Through hardware-level processing and redundant design, the problems of stuttering and packet loss in video transmission are solved, providing stable data input, reducing the computing power consumption of the edge AI computing box, ensuring the real-time performance and accuracy of monitoring, and improving the reliability and integrity of the system.
Smart Images

Figure CN224959686U_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of data transmission and processing, and specifically to a video surveillance device based on edge computing. Background Technology
[0002] In scenarios such as park security, video surveillance systems with a "front-end acquisition - edge processing - cloud analysis and backup" architecture are widely used. Video data collected by multiple cameras is transmitted to a local edge AI computing box for real-time analysis (such as personnel and fire identification) and to a remote cloud server for storage and analysis, forming a dual guarantee.
[0003] However, the existing system has significant drawbacks: video transmission is susceptible to electromagnetic interference and link issues within the park, resulting in stuttering, packet loss, and even data loss. This leads to a lack of stable data input for the edge AI computing boxes, causing recognition interruptions, monitoring blind spots, and an inability to promptly capture anomalies. Furthermore, the system lacks hardware-level preprocessing and redundancy mechanisms: fluctuations in the raw video stream bitrate and disordered frame order not only increase the transmission burden but also consume the computing power of edge devices; there is no local caching and retransmission method when transmission is interrupted, resulting in data loss and incomplete monitoring, making it difficult to meet security traceability requirements.
[0004] Therefore, improving transmission smoothness, reducing data problems, and ensuring the accuracy of identification and analysis have become urgent issues to be addressed. Utility Model Content
[0005] In view of the above problems, this application provides a video surveillance device based on edge computing to solve the technical problem that data is prone to lag and loss in the edge processing of video data, making it difficult to meet the needs of security traceability.
[0006] To achieve the above objectives, this application provides a video surveillance device based on edge computing, including multiple campus cameras, an edge-side video data preprocessing relay device, an edge AI computing box, and a cloud server;
[0007] The signal output terminals of the multiple campus cameras are respectively connected to the video input interface of the edge-side video data preprocessing relay device, and are used to transmit the raw video stream to the relay device;
[0008] The edge-side video data preprocessing relay device has a built-in hardware-level video codec chip and a main control chip. The signal output terminal of the video input interface is connected to the first signal input terminal of the main control chip. The hardware-level video codec chip is bidirectionally connected to the main control chip through a high-speed data bus for hardware-level frame synchronization calibration and adaptive bitrate adjustment of the original video stream.
[0009] The relay device has two video output interfaces: one connected to the video access port of the edge AI computing box, and the other connected to the remote receiving port of the cloud server, so as to synchronously transmit the optimized video stream processed by the codec chip to the edge AI computing box and the cloud server.
[0010] Furthermore, the edge-side video data preprocessing relay device also includes a dual-redundant storage module. The dual-redundant storage module consists of two independent storage chips. Both storage chips are bidirectionally connected to the main control chip through a data interface to form a main storage path and a backup storage path. This is used to synchronously mirror the video stream data processed by the hardware-level video codec chip and to retransmit the stored data during the interruption to the edge AI computing box and cloud server after the transmission link between the relay device and the edge AI computing box and cloud server is restored.
[0011] Furthermore, the hardware-level video codec chip is a Hisilicon Hi3519 series chip, and the high-speed data bus is a PCIe 3.0 bus or a GMII interface.
[0012] Furthermore, the video input and video output interfaces of the relay device are both gigabit RJ45 Ethernet ports or SFP optical ports, supporting the access and output of 8, 16, or 192 channels of video data, matching the multi-channel configuration parameters of the edge AI computing box.
[0013] Furthermore, the storage chip is an eMMC chip or a small SSD, and the data interface is an SPI bus interface or an eMMC interface.
[0014] Furthermore, the main control chip has built-in redundancy detection logic to periodically check the CRC value of the video stream data in the two storage chips to ensure that the storage content of the main storage path and the backup storage path is consistent.
[0015] Furthermore, the output buffer of the hardware-level video codec chip is connected to the main control chip via a high-speed data bus, and the processed optimized video stream is forwarded to the dual-channel video output interface of the relay device through the main control chip.
[0016] Furthermore, the edge AI computing box has a built-in computing chip and AI algorithm library for real-time identification, monitoring and analysis of the optimized video stream transmitted by the relay device; the cloud server is used for remote storage and background analysis processing of the optimized video stream.
[0017] Unlike existing technologies, the above technical solution is a video surveillance device based on edge computing, including an edge-side video data preprocessing relay device, an edge AI computing box, and a cloud server. The edge-side video data preprocessing relay device integrates a hardware-level video codec chip and a main control chip. The signal output terminal of the video input interface is connected to the first signal input terminal of the main control chip. The hardware-level video codec chip communicates bidirectionally with the main control chip via a high-speed data bus, used for hardware-level frame synchronization calibration and adaptive bitrate adjustment of the original video stream. The relay device's video output interface is configured with two paths: one connected to the video access port of the edge AI computing box, and the other connected to the remote receiving port of the cloud server, to synchronously transmit the optimized video stream processed by the codec chip to both the edge AI computing box and the cloud server. This technical solution adds an edge-side video data preprocessing relay device, effectively solving the problems of stuttering and packet loss in the transmission of the original video stream. Hardware-level codec chips perform frame synchronization calibration and adaptive bitrate adjustment on the video stream, optimizing data transmission quality, providing stable input for the edge AI computing box, ensuring the accuracy of its real-time recognition and monitoring analysis, and reducing recognition interruptions and monitoring blind spots caused by data issues. Simultaneously, the relay equipment shifts the video preprocessing stage to the hardware level, reducing the computing power consumption of the edge AI computing box and meeting its high-efficiency processing requirements. Furthermore, the optimized video stream is synchronously transmitted to the cloud server, ensuring the integrity of data stored and analyzed in the cloud and improving the overall system's monitoring reliability.
[0018] The above description of the utility model is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description
[0019] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of this application and other related content, and should not be considered as limitations on this application.
[0020] In the accompanying drawings of the instruction manual:
[0021] Figure 1 This is a schematic diagram of the device connection of the edge computing-based video surveillance device described in a specific embodiment;
[0022] Figure 2 This is a schematic diagram illustrating the composition and connection of the edge computing-based video surveillance device described in a specific embodiment;
[0023] Figure 3This is a block diagram of the specific modules of the relay device described in the specific implementation method; Detailed Implementation
[0024] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.
[0025] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.
[0026] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.
[0027] In the description of this application, the term "and / or" is used to describe the logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and A and B exist simultaneously. Additionally, the character " / " in this document generally indicates that the preceding and following objects have an "or" logical relationship.
[0028] In this application, terms such as “first” and “second” are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual quantity, hierarchy or order relationship between these entities or operations.
[0029] Unless otherwise specified, the use of terms such as “comprising,” “including,” “having,” or other similar expressions in this application is intended to cover non-exclusive inclusion, which does not exclude the presence of additional elements in a process, method, or product that includes the stated elements, such that a process, method, or product that includes a list of elements may include not only those defined elements but also other elements not expressly listed, or elements inherent to such a process, method, or product.
[0030] As understood in the Examination Guidelines, in this application, expressions such as "greater than," "less than," and "exceeding" are understood to exclude the stated number; expressions such as "above," "below," and "within" are understood to include the stated number. Furthermore, in the description of the embodiments in this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise explicitly specified.
[0031] In the description of the embodiments of this application, the space-related expressions used, such as "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "vertical," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," indicate the orientation or positional relationship based on the orientation or positional relationship shown in the specific embodiments or drawings. They are only for the purpose of describing the specific embodiments of this application or for the reader's understanding, and do not indicate or imply that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.
[0032] Unless otherwise expressly specified or limited, the terms "installation," "connection," "linking," "fixing," and "setting," as used in the description of the embodiments of this application, should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral setting; it can be a mechanical connection, an electrical connection, or a communication connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two components or the interaction between two components. For those skilled in the art to which this application pertains, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.
[0033] Please see Figures 1 to 3 This embodiment discloses a video surveillance device based on edge computing, suitable for park security scenarios. For example... Figure 1 As shown, the overall architecture of the edge computing-based video surveillance device includes multiple campus cameras, an edge-side video data preprocessing relay device (hereinafter referred to as "relay device"), an edge AI computing box, and a cloud server. Each component achieves stable connection and data interaction through hardware interfaces. The specific structure and working principle are as follows:
[0034] Among them, the multi-channel park cameras are selected as network cameras that support 1080P resolution. Their signal output ends are connected one-to-one to the video input interface of the relay device through Cat 5e network cables. In this embodiment, the video input interface and video output interface of the relay device are both gigabit RJ45 network ports, with a total of 16 access ports and 2 output ports. It can match the 16-channel video access configuration parameters of the edge AI computing box to meet the monitoring coverage needs of medium and large parks.
[0035] like Figure 2 and Figure 3 As shown, the relay equipment integrates a main control chip, a hardware-level video codec chip, and a dual-channel redundant storage module. The main control chip uses an ARM Cortex-A9 architecture chip, which is responsible for data forwarding and logic control functions. The hardware-level video codec chip is specifically a Hisilicon Hi3519 series chip, which communicates bidirectionally with the main control chip through a PCIe 3.0 high-speed data bus. When multiple campus cameras transmit raw video streams to the video input interface of the relay equipment, the video data is first forwarded by the interface controller to the main control chip, and then the main control chip pushes the raw video stream to the hardware-level video codec chip. This codec chip can perform two core hardware-level processing functions on the raw video stream: first, frame synchronization calibration, which removes duplicate frames and completes incomplete frames caused by link interference through hardware logic to ensure the continuous and orderly sequence of video frames; second, adaptive bitrate adjustment, which automatically stabilizes the bitrate in the 2-4Mbps range according to the video image complexity of different areas of the campus (such as densely populated passageways and power distribution rooms where equipment is stationary), avoiding transmission congestion caused by high bitrates while ensuring that the image clarity meets the recognition requirements.
[0036] The relay device's two video output interfaces are connected via network cables to the video access port of the edge AI computing box and the remote receiving port of the cloud server, respectively. The video stream, optimized by a hardware-level codec chip, is first transmitted back to the main control chip, and then synchronously forwarded by the main control chip to the two output interfaces, achieving dual-path transmission of "one path for local real-time processing and one path for cloud backup and analysis." The edge AI computing box has a built-in 8-core ARMA53 architecture computing chip and various AI algorithm libraries. The optimized video stream provides stable data input for it and can be directly used for real-time monitoring tasks such as personnel intrusion identification and open flame detection. It does not require additional computing power to sort out disordered data, greatly reducing the probability of identification interruption and monitoring blind spots. The cloud server receives and stores the optimized video stream, providing complete data support for later security incident tracing and big data analysis.
[0037] Meanwhile, the dual-redundant storage module within the relay device consists of two 32GB eMMC chips. Both eMMC chips communicate bidirectionally with the main control chip via eMMC interfaces, forming a primary storage path and a backup storage path. During video stream transmission, the main control chip synchronously mirrors and stores the video stream processed by the codec chip into the two eMMC chips at a rate of 1 second per slice. The main control chip has built-in redundancy detection logic, which checks the CRC value of the video slices in both chips every 500ms to ensure complete consistency between the stored contents. When the transmission link between the relay device and the edge AI computing box or cloud server is briefly interrupted due to loose network cables or electromagnetic interference, the dual-redundant storage module temporarily retains the video slices from the interrupted period. Once the link is restored, the main control chip automatically retrieves the video slices from the storage module and retransmits the data to the edge AI computing box and cloud server, avoiding monitoring gaps caused by data loss and further improving the integrity and reliability of the system monitoring.
[0038] In summary, this device, through hardware-level processing and dual-path redundancy design of the relay equipment, solves the problems of stuttering, packet loss, and data loss in the transmission of raw video streams, reduces the computing power consumption of the edge AI computing box, and ensures the integrity of cloud data, effectively improving the real-time performance, accuracy, and reliability of park security monitoring. Furthermore, in this embodiment, the connection method of "video input interface → main control chip → hardware-level video codec chip" significantly improves the collaborative processing efficiency and system stability of multiple video streams. Specifically, the main control chip, as the data scheduling core, first receives multiple raw video streams transmitted from the video input interface, completes interface protocol conversion (such as converting network data packets output by the camera into a format compatible with the codec chip), and then orderly distributes them to the hardware-level video codec chip via a high-speed data bus, avoiding signal conflicts and logical confusion that may occur when the codec chip directly connects to multiple interfaces. Meanwhile, the bidirectional communication between the main control chip and the codec chip includes both efficient transmission of video data (raw data input and optimized data feedback) and real-time interaction of control commands (such as dynamically adjusting frame synchronization parameters and bitrate thresholds), allowing the main control chip to monitor and regulate the processing status of the codec chip in real time. This architecture is particularly suitable for multi-channel concurrent video scenarios. The main control chip can coordinate the collaborative work of the codec chip, dual-channel redundant storage modules, and output interfaces through time-division multiplexing or parallel scheduling mechanisms, ensuring seamless connection of processing, storage, and transmission links for each video stream. This further reduces the risk of stuttering caused by data scheduling disorder and provides continuous and stable video data input for edge AI computing boxes and cloud servers.
[0039] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.
Claims
1. An edge computing based video monitoring device, characterized by, The application relates to a video data pre-processing relay device for edge AI computing box and cloud server. The signal output end of the multi-path park camera is connected with the video input interface of the edge-side video data pre-processing relay device, and is used for transmitting original video stream to the relay device. The edge-side video data pre-processing relay device is internally provided with a hardware-level video coding chip and a main control chip, the signal output end of the video input interface is connected with the first signal input end of the main control chip, and the hardware-level video coding chip is bidirectionally connected with the main control chip through a high-speed data bus, and is used for carrying out hardware-level frame synchronization calibration and self-adaptive code rate adjustment on the original video stream. The video output interface of the relay device is provided in two paths, one path is connected with the video access port of the edge AI computing box, and the other path is connected with the remote receiving port of the cloud server, so that the optimized video stream processed by the coding chip is synchronously transmitted to the edge AI computing box and the cloud server.
2. The edge computing based video surveillance apparatus according to claim 1, characterized in that, The edge-side video data pre-processing relay device further comprises a dual-path redundant storage module, the dual-path redundant storage module is composed of two independent storage chips, the two storage chips are bidirectionally connected with the main control chip through a data interface, form a main storage path and a standby storage path, are used for synchronously mirroring and storing the video stream data processed by the hardware-level video coding chip, and after the transmission link between the relay device and the edge AI computing box and the cloud server is recovered, the storage data during the interruption is supplemented to the edge AI computing box and the cloud server. 3.The edge computing based video surveillance device according to claim 1, characterized in that, The hardware-level video coding chip is a Hi3519 series chip of HiSilicon, and the high-speed data bus is a PCIe 3.0 bus or a GMII interface.
4. The edge computing based video surveillance apparatus according to claim 1, wherein, The video input interface and the video output interface of the relay device are both gigabit RJ45 network interfaces or SFP optical interfaces, can support the access and output of 8-path, 16-path or 192-path video data, and match the multi-path configuration parameters of the edge AI computing box.
5. The edge computing based video surveillance device of claim 2, wherein, The storage chip is an EMMC chip or a small SSD, and the data interface is an SPI bus interface or an eMMC interface.
6. The edge computing based video surveillance device of claim 2, wherein, The main control chip is internally provided with a redundancy detection logic, is used for periodically checking the CRC values of the video stream data in the two storage chips, and ensures that the storage contents of the main storage path and the standby storage path are consistent. 7.The edge computing based video surveillance device according to claim 1, characterized in that, The output buffer area of the hardware-level video coding chip is connected with the main control chip through the high-speed data bus, and the optimized video stream processed by the main control chip is forwarded to the dual-path video output interface of the relay device. 8.The edge computing based video surveillance device of claim 1, wherein, The edge AI computing box is internally provided with a computing power chip and an AI algorithm library, is used for carrying out real-time identification and monitoring analysis on the optimized video stream transmitted by the relay device, and the cloud server is used for remotely storing and background analyzing and processing the optimized video stream.
