Human and Face Detection
Detects people, faces and body position to understand where participants are in the meeting space.
Make cameras understand the meeting room
Bluesource-Tech AI imaging algorithms combine human detection, scene understanding and audio-video linkage to keep every meeting frame compact, natural and focused on the right people.
Panorama
Best View
Voice Tracking
Smart Gallery
From raw image to meeting-aware composition
The imaging engine recognizes people, counts participants, tracks motion and connects with audio direction so the output frame follows the actual conversation.
Scene Understanding
Evaluates room boundaries, seating layout and participant changes to keep framing stable and useful.
Audio-Video Linkage
Uses audio DOA together with image detection to identify the current speaker and switch focus more naturally.
Multi-View Output
Builds panorama, close-up and gallery views for room systems, meeting cameras and video endpoints.
AI Imaging Modes
Four modes cover full-room awareness, dynamic composition, speaker focus and multi-person gallery output.
01
Panorama Mode
Keeps the full room visible for fixed layouts, whiteboard sharing, live demonstrations and multi-person discussion.
- Full spatial context
- Stable room overview
- Suitable for fixed meetings
Everyone in frame
02
Best View Mode
When one participant leaves, the camera automatically recomposes around the remaining people instead of leaving empty space in frame.
- Auto detects headcount and edge changes
- Reframes from four people to three people
- Reduces empty wall and table area
Frame recomposed
03
Voice Tracking Mode
Combines sound direction and human detection to lock the frame on the active speaker.
- Audio DOA and video detection
- Speaker-centered framing
- Useful for hybrid meetings and training
Speaker locked
04
Smart Gallery Mode
Splits multiple participants into close-up windows so remote attendees can see everyone more clearly.
- Independent participant windows
- Near-view reconstruction
- Better remote visibility
Gallery output
Image Processing Chain
Camera Frames
Captures high-resolution video from USB cameras, embedded modules or integrated video bars.
People and Scene
Recognizes people, faces, room edges and participant movement.
Mode Selection
Applies panorama, best view, voice tracking or smart gallery according to the meeting scenario.
Meeting Frame
Provides a composed video stream for conferencing platforms and collaboration terminals.
Bring meeting-aware imaging into your product
Bluesource-Tech can adapt AI imaging modes to conference cameras, audio-video modules and all-in-one video endpoints.