Image Annotation for Computer Vision

  • Semantic segmentation
    Create project - > Upload images - >Create attributes - >
    {
    "name": "Semantic Segmentation street view driving",
    "desc": "Scene recognition and detection of road, driving lanes, cars and other vehicles, pedestrians",
    "task_type_id": 2,
    "attributes": {road, emergency lane, driving lane, buildings, traffic light, car, car windshield, car plate, bike, bycicle, pedestrian }
    }
    FORMATEXPLORE
 
  • Bounding boxes
    Create project - > Upload images - >Create attributes - >

    {
    "name": "Airport - object Detection",
    "desc": "Object recognition - at airports",
    "task_type_id": 1,
    "attributes": {airplane}
    }
    FORMATEXPLORE
 
  • Polygonal annotation
    Create project - > Upload images - >
    Create attributes - >
    {
    "name": "Aerial camera: port freight size",
    "desc": "Ships and containers size annotation",
    "task_type_id": 2,
    "attributes": {container, boat, ship}
    }
    FORMATEXPLORE
 
  • Polyline annotation
    Create project - Upload images - Create attributes -

    {
    name: "Driving line detection",
    describe Building the grid of driving line annotation,
    task_type_id 4,
    attributes {driving lane, emergency lane, lane border}
    }
    FORMATEXPLORE
 
  • Geo-tagging structures
    Create project - > Upload images - >
    Create attributes - >

    {
    "name": "Aerial camera: infrastructure detection",
    "desc": "infrastructure size and locations",
    "task_type_id": 4,
    "attributes": {container, commercial building, warehouse}
    }
    FORMATEXPLORE
 

If you are interested in a full business data set solution using satellite and aerial imagery, explore Alternative Data sets

How it Works
 
We work on complex AI image solutions to companies developing Industrial Automation,   Anomaly Detection and Autonomous Driving companies and using extensively data collected from sensors, cameras to improve their operations and offer AI-solutions.
It starts from understanding the requirements for the training sets, understanding the type of labeling, and size of desired dataset.
Semantic Segmentation
Robo portal
3edtd