Hardware/feature/liquid lense

[Yasserelhaddar]

Liquid Lens & Autofocus Control

Summary

Adds hardware-level liquid lens and autofocus control for Basler cameras with connected Optotune ECC controllers. The feature spans all architecture layers — backend, core wrappers, service endpoints, and documentation — following the existing optional-capability pattern.

Changes by Layer

Backend (cameras/backends/)

• 7 optional methods added to CameraBackend base class (log + raise NotImplementedError)
• Full Basler implementation: auto-detects lens via LensConnection GenICam node
• get_lens_status() returns a data dict (no exception) even for cameras without a lens
• Optical power get/set with range validation, one-shot autofocus with accuracy selection and polling
• Focus config mapping to GenICam nodes (ROI, stepper, edge detection, source)
• export_config / import_config extended with lens parameters

Core (cameras/core/)

• 7 async methods on AsyncCamera (lock-protected for setters)
• 7 sync wrappers on Camera
• configure() accepts optical_power; supports_feature reports liquid_lens

Service (services/cameras/)

• 6 REST endpoints under cameras/lens/ and cameras/focus/
• Request models: OpticalPowerRequest, TriggerAutofocusRequest, FocusConfigRequest
• Response models: LensStatus, LensStatusResponse
• TaskSchemas for MCP tool exposure
• /cameras/capabilities now includes supports_liquid_lens

Documentation

• Camera service README: focus control endpoints, MCP tools, configuration parameters
• Hardware README: usage example, auto-detection notes

How It Works

from mindtrace.hardware.cameras.core.camera import Camera

camera = Camera(name="Basler:ace2_001")

# Auto-detection — no exception for unsupported cameras
status = camera.get_lens_status()
if status["connected"]:
    # Manual focus control
    camera.set_optical_power(5.0)  # diopters

    # One-shot autofocus (contrast-based)
    camera.trigger_autofocus(accuracy="Accurate")

    # Fine-tune AF behavior
    camera.set_focus_config(roi_size="Size64", edge_detection=True)

camera.close()

Via REST:

curl -X POST http://localhost:8002/cameras/focus/autofocus \
  -H "Content-Type: application/json" \
  -d '{"camera": "Basler:ace2_001", "accuracy": "Normal"}'