Browse categories
Explore
Fiverr Pro
English
$
USD
I will build multi camera ai tracking for mask detection and motion analytics
samad
About this gig
Stop monitoring manually. Start automating safety.
In high-stakes environmentshospitals, construction sites, and manufacturing plantscompliance isn't optional. I build Industrial-Grade AI Video Analytics that combine real-time Face Mask Detection with advanced Person Movement Tracking to ensure 24/7 safety oversight without human error.
The Synergy: Why Both Matter
Most developers offer one or the other. I integrate them into a single, high-performance pipeline:
- Compliance Monitoring: Instant detection of PPE/Face Mask violations with timestamped logging.
- Behavioral Tracking: Beyond simple detection, I track individual movement paths to identify "High-Risk" behaviors or unauthorized entry into restricted zones.
- RTSP Scalability: My systems don't just work on one webcam; they are optimized to handle multi-camera RTSP feeds (16+) with zero lag.
Key Features of the System:
- Dual-Stream Intelligence: Real-time Mask/No-Mask classification paired with unique Person IDs (Re-ID).
- Zone-Aware Analytics: Define specific "Mask-Mandatory Zones" vs. "Common Areas" to reduce false alerts.
- Motion & Velocity Insights: Track if a person is running, loitering, or entering a hazardous area without prop
Get to know samad
samad
AI MLOps Engineer
- FromBangladesh
- Member sinceJan 2026
- Avg. response time1 hour
- Last delivery3 weeks
Languages
Bengali, English
If your company needs to analyze RTSP feeds, detect abnormal events, or deploy scalable AI solutions to the edge or cloud, I can take you from concept to a production-ready system that delivers measurable impact. With 5+ years of hands-on experience and certifications from IBM, Google, Meta, and Microsoft, I specialize in designing intelligent systems that combine Computer Vision, Deep Learning, and LLM- powered automation to solve complex business problems efficiently.
My Portfolio
FAQ
Can this system distinguish between a real mask and a hand covering the face?
Yes. By fine-tuning the YOLOv8/v11 models on high-quality industrial datasets, the system is trained to identify the specific textures and shapes of surgical, N95, and cloth masks. This significantly reduces false positives compared to basic "skin-color" detection scripts.
Will I need an expensive GPU to run this on 16+ cameras?
Not necessarily. While a dedicated GPU (like an RTX series) is ideal for servers, I specialize in Edge AI Optimization. I can use TensorRT and Quantization to make the models run at high FPS on affordable hardware like the NVIDIA Jetson series or even optimized Intel CPUs using OpenVINO.
How does the system track people across different camera views?
I implement Multi-Object Multi-Camera Tracking (MOMCT). The system assigns a unique "Persona ID" to an individual based on visual features. If they move from Camera A to Camera B, the system attempts to maintain that ID, allowing you to track their entire movement path across your facility.
Is the data private? Does it upload my video to the cloud?
That is entirely up to you. I can build a 100% On-Premise solution where all AI processing happens locally on your hardware. This ensures maximum data privacy and zero monthly cloud subscription fees. If you prefer a cloud-based dashboard, I use secure encrypted APIs to sync only the metadata (analy
Can I set "Mask-Mandatory" zones and "Mask-Free" zones?
Absolutely. I implement Polygon Zone Awareness. You can draw specific areas on your camera feed (like a sterile lab or a construction entrance) where the mask alert is active, while ignoring people in "Break Rooms" or "Outdoor Zones" where masks may not be required.
What happens if a person is running or moving too fast for the camera?
My Motion Analytics engine calculates velocity in real-time. If a person is moving at a speed that suggests "Running," the system can trigger a specific "Safety Alert." By using high-speed trackers like ByteTrack or BoT-SORT, we maintain high accuracy even with fast-moving subjects.
