Search

CN-121985148-A - Remote management method and system for intelligent video analysis algorithm rules

CN121985148ACN 121985148 ACN121985148 ACN 121985148ACN-121985148-A

Abstract

The invention relates to the technical field of video monitoring, in particular to a remote management method and a remote management system of intelligent video analysis algorithm rules, wherein the method comprises the steps of respectively deploying a group of algorithm rules for analyzing monitoring videos aiming at different management and control areas at an edge computing end; the method comprises the steps of providing a remote server, providing a remote server with a heartbeat signal, providing an algorithm rule setting condition by the edge computing end, providing an updating instruction to the remote server by the edge computing end, providing a corresponding operation by the remote server according to the remote instruction, feeding back an execution result to the remote server by the edge computing end, primarily identifying whether an abnormality exists in a monitoring video by the algorithm rule at the edge computing end, outputting an alarm picture to the remote server, secondarily identifying the alarm picture by an AI intelligent agent at the remote server, and generating a reservation or discarding suggestion for the alarm picture. The method can improve the applicability of the intelligent video algorithm in industrial application and the accuracy of early warning of abnormal events.

Inventors

  • HE BING

Assignees

  • 东信朗云(北京)科技发展有限公司

Dates

Publication Date
20260505
Application Date
20260210

Claims (10)

  1. 1. The remote management method of the intelligent video analysis algorithm rule is characterized by comprising the following steps of: respectively deploying a group of algorithm rules for analyzing the monitoring video aiming at different control areas at an edge computing end; the edge computing end submits heartbeat signals and algorithm rule setting conditions of each management and control area to the remote server end; The remote server side sends corresponding updating instructions to the edge computing side according to the received related data of different control areas; The edge computing end executes corresponding operation according to the received remote instruction and feeds back an execution result to the remote server; The method comprises the steps that whether an abnormality exists in a monitoring video is primarily identified at an edge computing end through an algorithm rule, an alarm picture is output, and the alarm picture is sent to a remote server end; and deploying an AI intelligent agent at the remote server, and carrying out secondary identification on the alarm picture through the AI intelligent agent to generate a reservation or discarding suggestion aiming at the alarm picture.
  2. 2. The remote management method of intelligent video analysis algorithm rules according to claim 1, wherein a plurality of management areas are divided on an edge computing end interface, each management area corresponds to a group of algorithm rules, each algorithm rule is provided with a unique ID, each management area comprises four buttons, namely a setting rule button, a pause rule button, a resume rule button and a delete rule button, wherein the setting rule button is used for configuring corresponding algorithm rules for the management areas, the pause rule button is used for temporarily closing all algorithm rules configured in the currently selected management area but not deleting data, the resume rule button is used for enabling configuration of history storage, and the delete rule button is used for executing deletion operation on the configured corresponding algorithm rules.
  3. 3. The remote management method of the intelligent video analysis algorithm rule according to claim 1, wherein a heartbeat signal submitted by the edge computing end to the remote server end is used for indicating whether monitoring equipment of each control area is online or not, algorithm rule setting conditions comprise whether each control area is provided with algorithm rules, all algorithm rule IDs, in-use algorithm rules and suspended algorithm rules, and the remote server end displays online conditions, algorithm rule lists and running states of the monitoring equipment in each control area in a management interface.
  4. 4. The remote management method of intelligent video analysis algorithm rules according to claim 1, wherein the remote server sends a rule suspension instruction or a rule restoration instruction to the edge computing end according to the algorithm rule ID; And after the edge computing end receives the rule recovery instruction, restarting the previously stored algorithm rule to perform video analysis on the corresponding control area.
  5. 5. The remote management method of intelligent video analysis algorithm rules according to claim 1, wherein when the remote server sends the update instruction to the edge computing end, check bit authentication information is synchronously sent, when the edge computing end receives the update instruction and fails to pass the check bit authentication, the current update instruction is discarded, otherwise, the current update instruction is executed.
  6. 6. The remote management method of intelligent video analysis algorithm rules according to claim 1, wherein an operation log of the algorithm rules is recorded at an edge computing end, and a time stamp of each time of setting, suspending and resuming operation of the algorithm rules, an operation source of local or remote, an instruction check bit authentication information abstract and an operation result state are recorded in the operation log.
  7. 7. The remote management method of intelligent video analysis algorithm rules according to claim 1, wherein the edge computing device determines that the warning is valid after receiving the reservation advice of the AI intelligent agent for the current warning picture, and determines that the false warning exists after receiving the discard advice for the current warning picture, and filters out the current warning picture.
  8. 8. The remote management method of intelligent video analysis algorithm rules according to claim 1, wherein the alarm pictures suggested to be kept and the alarm pictures suggested to be discarded are continuously stored in an alarm data pool and labeled, and the algorithm rules are continuously trained and optimized based on the alarm pictures in the alarm data pool.
  9. 9. A remote management system for intelligent video analysis algorithm rules, characterized in that the method of any one of claims 1-8 is adopted, and the system comprises an edge computing end and a remote service end; An intelligent video analysis module is deployed on the edge computing end and comprises a rule execution unit, a reporting unit and a first alarm unit; the remote server is provided with a monitoring unit and a second alarm unit; the rule execution unit deploys a group of algorithm rules for analyzing the monitoring video aiming at different control areas respectively; the reporting unit submits heartbeat signals and algorithm rule setting conditions of each control area to the monitoring unit; the monitoring unit sends corresponding updating instructions to the rule execution unit according to the received related data of different control areas; The rule execution unit executes corresponding operation according to the received remote instruction and feeds back an execution result to the monitoring unit; The first alarm unit primarily identifies whether the monitoring video has abnormality or not through algorithm rules, outputs an alarm picture and sends the alarm picture to the second alarm unit; The second alarm unit carries out secondary identification on the alarm picture through the AI intelligent agent, and generates a reservation or discarding suggestion aiming at the alarm picture.
  10. 10. The remote management system of intelligent video analysis algorithm rules according to claim 9, wherein the intelligent video analysis module further comprises an optimizing unit, the optimizing unit continuously stores the alarm pictures recommended to be kept and the alarm pictures recommended to be discarded in the alarm data pool, tags the alarm data pool, and continuously trains and optimizes the algorithm rules based on the alarm pictures in the alarm data pool.

Description

Remote management method and system for intelligent video analysis algorithm rules Technical Field The invention relates to the technical field of video monitoring, in particular to a remote management method and a remote management system for intelligent video analysis algorithm rules. Background Currently, video monitoring systems have been applied to various industries, in which a large number of high-definition cameras are installed and deployed by many enterprises, and the large amount of data presents a great challenge to video monitoring due to the limited number of personnel in security and security departments. In the face of real-time video data which is often taken as a unit of T, an abnormal video is not easy to find in time by adopting a manual monitoring mode, and the searching of an abnormal event from stored historical video data is difficult. Because the intelligent video analysis algorithm can effectively extract abnormal frames in real-time or historical video data according to a pre-trained rule, the intelligent video analysis algorithm can effectively monitor the video data instead of manpower, find the abnormal data in the video data, provide an alarm, and greatly improve the analysis efficiency of the video monitoring data. However, in practical application, the method has the problems that the setting and canceling of the intelligent video algorithm rule is manually completed, a large algorithm rule maintenance workload is formed, the consistency between the reset algorithm rule and the algorithm rule before canceling is poor, the intelligent video analysis algorithm generates a high false alarm rate along with the continuous change of a scene, the alarm formed by a large amount of false alarm data forms an alarm elimination working pressure, and the trust degree of a user on the algorithm alarm is reduced. Therefore, how to improve the applicability of the intelligent video algorithm in industrial application and the accuracy of early warning of abnormal events is a problem that needs to be solved by those skilled in the art. Disclosure of Invention In view of the above, the present invention provides a remote management method and system for intelligent video analysis algorithm rules, so as to overcome or at least partially solve the above problems. In order to achieve the above purpose, the present invention adopts the following technical scheme: In a first aspect, the present invention provides a remote management method for rules of an intelligent video analysis algorithm, including the following steps: respectively deploying a group of algorithm rules for analyzing the monitoring video aiming at different control areas at an edge computing end; the edge computing end submits heartbeat signals and algorithm rule setting conditions of each management and control area to the remote server end; The remote server side sends corresponding updating instructions to the edge computing side according to the received related data of different control areas; The edge computing end executes corresponding operation according to the received remote instruction and feeds back an execution result to the remote server; The method comprises the steps that whether an abnormality exists in a monitoring video is primarily identified at an edge computing end through an algorithm rule, an alarm picture is output, and the alarm picture is sent to a remote server end; and deploying an AI intelligent agent at the remote server, and carrying out secondary identification on the alarm picture through the AI intelligent agent to generate a reservation or discarding suggestion aiming at the alarm picture. Further, a plurality of control areas are divided on an edge computing end interface, each control area corresponds to a group of algorithm rules, each algorithm rule is provided with a unique ID, each control area comprises four buttons, namely a setting rule button, a pause rule button, a resume rule button and a delete rule button, wherein the setting rule button is used for configuring corresponding algorithm rules for the control area, the pause rule button is used for temporarily closing all the algorithm rules configured in the currently selected control area but not deleting data, the resume rule button is used for enabling configuration stored in history, and the delete rule button is used for executing delete operation on the configured corresponding algorithm rules. Furthermore, the heartbeat signal submitted by the edge computing end to the remote server end is used for indicating whether the monitoring equipment of each management and control area is on line or not, the algorithm rule setting condition comprises whether each management and control area is provided with an algorithm rule, all algorithm rule IDs, the algorithm rule in use and a suspended algorithm rule, and the remote server end displays the on-line condition, the algorithm rule list and the running state of the monitoring equipme