# Create & deliver custom video analytics solutions in minutes Lumeo is a “no-code” custom video analytics platform that allows providers to deliver new solutions easily with drag & drop tools and pre-built analytic building blocks. ## Analytics Library ### Start with our ever-growing library of ready-to-use AI models or bring your own. - **Bodypose 2D**: Detects human body poses in video streams. - **People Detection (Crowds)**: Detects heads for counting in crowded environments. - **Fire Detection**: Detects possible fire or smoke in video frames. - **Forklift Detection**: Detects forklifts in warehouse environments. - **Person Attribute Detection**: Detects clothing colors and attributes on persons. - **Hands Detection**: Detects hands and fingers in video frames. - **Hardhat Detection**: Detects if workers are wearing hardhats. - **Face Mask Detection**: Detects if people are wearing face masks. - **Barcode Detection**: Detects and reads barcodes from video frames. - **Person Detection**: Detects people in video frames. - **Vehicle Detection**: Detects vehicles including cars, trucks, and buses. - **High Vantage Vehicle Detection**: Vehicle detection optimized for high-angle cameras. - **Vehicle Detection (Low Vantage)**: Vehicle detection from street-level cameras. - **Vehicle Type Classification**: Identifies the type of vehicles in video frames. - **Large Vision Models**: Gemini, GPT-4o, Claude Vision, AWS Bedrock scene understanding. - **Weapon Detection**: Detects firearms and weapons for security applications. - **Emergency Vehicle Detection**: Identifies ambulances, fire trucks, and police vehicles. - **Adult vs Child Classification**: Classifies detected persons as adults or children. - **Generic Detector**: Detection model with text-based prompt grounding. - **Generic Classifier**: Classify objects or events using natural language prompts. - **Person ReIdentification**: Re-identify and track people across multiple cameras by appearance. - **Face ReIdentification**: Re-identify individuals across cameras using facial features. ## Go beyond the bounding box. Connect model detections and classifications to rules and third-party systems ### Rules - Proximity - Speed - Direction - Occupancy - Line crossing ### Integrations - Milestone - Genetec - Elasticsearch - BMS - Webhooks - HikCentral ### Video Utilities - Blur - Watermark - Stream RTSP - Stream RTMP - Stream WebRTC - Clips - Snapshots ### Recognitions - Face recognition - License plate recognition - Make / Model / Color recognition ### Video source - Use cameras or RTSP streams as video input ### Track objects - Detect tracked objects and assign a unique identifier to each new object ### Encode Video - Encode raw video frames and output an encoded stream ### Custom function - Manually process or modify frames and metadata ### Recognize barcodes - Recognize barcodes in video ### Add metadata - Inject a file's contents as metadata to each frame ### Blur Objects - Blurs any detected objects in the video ### Save Clip - Save a clip when a specific condition is met ### Demultiplex Video - Separate a multiplexed stream into its constituent streams ### Transform Video - Manipulate video transformations, resolution, and frame rate ### Trigger Genetec Alarm - Trigger an alarm in Genetec Security Center when a specific condition is met ### Stream RTMP - Create a custom output stream to send to YouTube, Mux, Wowza and more ### Detect motion - Detect motion in the input frames and add confidence levels as metadata ### Line crossing counter - Counts how many particular tracked objects crossed each line ### Trigger HikCentral Alarm - Trigger an alarm in HikCentral ### Filter Stream - Filter frames based on condition or interval ### Stream RTSP - Create an RTSP output stream for use in other pipelines or systems ### Occupancy monitor - Extracts queue, dwell and occupancy information based on how long detected objects have been present in a specified region ### Trigger Milestone Event - Trigger an analytics event in XProtect ### Webhook - Send metadata from the pipeline to a REST API at a periodic interval or based on a trigger ### Azure face recognition - Lookup detected faces using Azure Face Recognition API ### Filter objects - Filter objects based on their presence within or outside a region of interest ### Public to Elasticsearch - Publish metadata to Elasticsearch for reporting and dashboards ### Save Snapshot - Save a snapshot when a specific condition is met ### Display Stream Info - Display stream meta and info on the video stream ### Publish to Google Sheets - Publishes selected metadata using Google Sheets API ### Pose recognition - Parses Bodypose model output and augments metadata ### License plate parser - Extracts and post-processes license plates recognized using the License Plate Recognition - Edge AI Model ### Display Watermark - Display a watermark on the video stream ### AI Model - Run AI model inference on input frames ### Tile Video in Grid - Arrange the constituent streams of a multiplexed stream into a grid ### Stream WebRTC - Create a WebRTC output stream to use as a video stream in other pipelines or view it in a web browser ### Multiplex Video - Combine multiple input streams into one multiplexed stream for optimized processing ### License plate MMC cloud - High accuracy license plate + make/model/color recognition using Lumeo's cloud service ### License plate MMC recognizer - High accuracy license plate + make/model/color recognition using your own Triton/Docker Snapshot API key ### Proximity detector - Detect and create groups of objects when they are close to or far from each other ### Trigger Pushover Notification - Send a push notification using Pushover app ### Overlay Metadata - Draw model inference data on a video stream ### Speed estimation - Estimates speed of objects in a ROI and sets triggers based on low/high thresholds ### Trigger SMS Alert - Send an SMS alert when a specific condition is met ## Analytics Builder ### Deploy or Build any use case from anywhere, in record time. #### Production-ready templates - Pre-built templates - Deploy a wide range of use cases without lifting a finger. #### Flexible & customizable - Start with a template and change any aspect of it. Add new Rules, send alarms to a different VMS, and more. #### Drag & drop tools › #### APIs and custom code › ## Search & Dashboards ### AI-powered analytics at your fingertips. ## Plug & Play ### Get up and running fast with existing or new infrastructure. ### Any camera - Use any media source, for real-time analytics. Built-in support for ONVIF/USB Cameras, RTSP streams, HLS and MPEG Dash feeds. ### Any recorded video - Run analytics on video clips or recorded footage. Analyze & infer the fact or run a forensic analytic. ### VMS and NVRs - Integrates with VMS and NVRs for a single pane of glass. Extract streams from your VMS, process with Lumeo, and feed them back. ### Built-in monitoring - Deliver reliability easily, with real-time monitoring of your cameras and analytics. Alerts, webhooks, and integrations make it easy to reduce downtime. ## Edge or Cloud ### Design centrally, deploy anywhere: Lumeo cloud, Your cloud, On-prem GPUs. - Cloud-based console to design, deploy, and manage from anywhere. - Deploy analytics securely and privately. - Process in Lumeo cloud, Your cloud, or Edge (on-prem GPUs or NVIDIA Jetson). ## Hardware Partners ### Works on hardware you already own - Deploy Lumeo-Ready AI gateways or run on your existing GPU servers, edge devices, and cloud infrastructure.