API Reference

Version: v1.0.0

This API enables automatic camera calibration in Scenescape, supporting both AprilTag and markerless methods. It also supports sensor-agnostic perceptual sensor calibration, aligning a point cloud from any perceptual sensor (LiDAR, depth camera, stereo, photogrammetry) to a scene’s 3D model.

Base URL: https://localhost:8443/v1

Endpoints

Service Status

• GET /status — Check if the calibration service is running.

Scene Registration

• POST /scenes/{sceneId}/registration — Register a scene for calibration processing.

• GET /scenes/{sceneId}/registration — Get the status of scene registration.

• PATCH /scenes/{sceneId}/registration — Notify the service that a scene has been updated and needs re-processing.

Camera Calibration

• POST /cameras/{cameraId}/calibration — Start camera calibration by uploading an image and (optionally) camera intrinsics.

• GET /cameras/{cameraId}/calibration — Get the status and result of camera calibration, including pose and calibration data.

Perceptual Sensor Localization

• POST /perceptual-sensors/{sensorId}/localization — Localize a perceptual sensor by supplying a point cloud (base64-encoded; PCD by default, PLY also accepted) to align against a scene’s 3D model and compute the sensor-to-scene transform.

• GET /perceptual-sensors/{sensorId}/localization — Get the status and result of perceptual sensor localization, including the 4x4 transform, fitness, and inlier RMSE.

Schemas

The API uses structured request and response schemas, including:

• ServiceStatus

• SceneRegistrationTriggerResponse

• SceneRegistrationStatusResponse

• CameraCalibrationRequest

• CameraCalibrationTriggerResponse

• CameraCalibrationStatusResponse

• PerceptualSensorLocalizationRequest

• PerceptualSensorLocalizationTriggerResponse

• PerceptualSensorLocalizationStatusResponse

• Error

For full schema details and example payloads, see the OpenAPI YAML file below.

---