Nowadays, camera equipment are installed everywhere. On a typical day in a large city, you might see hundreds of surveillance cameras (CCTV). Such a large number of cameras creates a huge amount of data that security and surveillance personnel need to track every day. If we were to simply use human resources, this would be almost impossible.
Some systems have the ability to alert security guards of potential dangers from what the camera is recording. They can read car license plates and recognize certain faces. Some systems use specific algorithms to detect suspicious activities, such as when a suspicious group of people loiters near a door where the aisle is prohibited, or to detect some unnoticed baggage.
Inwhere, Briefcam's technology is outlined with a combination of Machine Learning and AI that allows video investigators to review video time in just a few minutes and quickly identify interested people and objects.
One of the key features of Briefcam software is that it helps aggregate videos that are several hours long in minutes; analyzes video quantitation to draw useful data from the subject's faces, colors, vehicles ... to help in investigating and making action decisions; and reacts immediately to important changes in the environment.
1. Video Analytics Solutions overview
Connection mechanism between components
Video Synopsis (VS) server
The server handles access interface, authorizes the users, handles the stream as well as stores the video.
Research Server (RS):
Analysis and reporting servers
Processing Server (PS):
Video processing server with multiple GPU cards dedicated to decoding, rendering, object extraction, and classification.
Components in Briefcam solutions
Video search application software includes real-time online analytics and offline search analysis.
- Summarize a video into a short video by simultaneously showing objects at different times.
- Search for objects (people and vehicles) based on basic features and information (gender, color, behavior) from the data source on the video CSDL stored on the camera management system (VMS) and independent video files.
- Object filtering feature: Filter by color, gender, direction of movement, regions, vehicles.
- Face detection feature.
Analyzing and issue warnings:
- Real-time analysis:
- Analysis based on pre-set schedule
User sets rules to monitor, analyze real-time video sources and alert when there is an event within these following rules:
- Virtual fence
- Intruders detection
- Crowd detection
- Detect prowling objects
- Movement detection
- Detect the direction of movement
The system's decentralized architecture allows flexible scaling according to demand. By adding functional servers, the performance is consistent with its sizing material.
Minimum 24-month support warranty period.
2. Technical specifications
Reference price: 3000 $ / channel
(PLEASE CONTACT US DIRECTLY FOR MORE ACCURATE CONSULTATION AND QUOTATION)