The video analysis box adopts a cross platform main approach to write underlying algorithms and upper level application software, which can theoretically adapt and support any operating system. Currently, the tested operating systems are Ubuntu 18.04 and Windows 7/8/10/12/16. In order to better adapt algorithms, improve query speed, and store big data, our system database adopts a self-developed database AIDB.
1、 Functional module real-time preview AI channel real-time preview
The client preview function interface supports real-time preview of AI node channel real-time videos, which have an analysis overlay effect. It supports pan tilt control, preset position management, image capture, double click image enlargement, full screen and other operations.
Real time preview of monitoring channels
The client preview function interface supports real-time preview of monitoring camera node channel real-time videos. This video does not have an analysis overlay effect and supports pan tilt control, preset position management, image capture, double click image enlargement, full screen and other operations.
View Management
In the client preview function interface, it supports saving opened videos as view templates, creating 9999 views, and setting the client to automatically load templates after startup.
Data Query
The client has a data query function interface, which supports selecting a certain channel, event type, and time range to query the alarm capture data generated by the system‘s intelligent capture channel. It also supports downloading the queried data (including images) to the client‘s local disk.
statistical analysis
The system has a massive statistical analysis module that automatically counts data for each event type, with the smallest statistical granularity in hours. It supports quick querying of statistical data by daily, monthly, annual, time period, channel, and other methods.
system configuration
In the system configuration function module, administrator users can operate and set the following functions:
Application rules for AI channel corresponding cameras and analysis recognition algorithms
Corresponding cameras for monitoring channels, and settings for recording, etc
Upload third-party platform HTTP protocol configuration
Linkage email sending configuration
System maintenance operations (system restart, time synchronization)
Secondary development
This system provides the following secondary development protocol interfaces:
The third-party HTTP request protocol is used by users to initiate HTTP post request response operations to the system. After receiving the command, the system executes the corresponding operations, mainly including: requesting a list of system channels, querying the capture event records of channels, pulling images, adding personnel (faces), querying personnel, deleting personnel, detecting faces, detecting objects, and other interfaces
The system actively uploads the capture event HTTP protocol, and the user adds a third-party receiving IP address and port information in the system configuration. The system uploads the capture event records in real time to this third-party IP address port. Please refer to the corresponding documentation for specific protocols
Real time video on demand and capture event real-time reception protocol, the system plays real-time videos (including analyzing overlay effects) through a plugin free web page, subscribes to receive alarm events, and details can be found in the corresponding DEMO program.
