cortenforge-vision-core 0.1.2

Core vision interfaces and capture/overlay helpers for the CortenForge stack. Renamed from vision_core; old crate deprecated—use this name.
Documentation

vision_core

crates.io docs.rs MSRV

Shared detector/capture/overlay interfaces for sims, tools, and inference.

Deprecated: this crate was renamed to cortenforge-vision-core. Please depend on the new crate name going forward.

Contents

  • interfaces: Frame/DetectionResult/Label/FrameRecord; Detector/FrameSource/Recorder traits.
  • overlay: box normalize + draw helpers.
  • capture: CaptureLimit (max frames).
  • prelude: re-exports interfaces, overlay helpers, CaptureLimit.

Usage

  1. Add vision_core as a dependency.
  2. Import via vision_core::prelude::* for interfaces/overlay helpers.
  3. Implement Detector/DetectorFactory in your crate; keep heavy backends (Burn) behind feature flags outside vision_core.
  4. Use BurnDetectorFactory only if you need a feature-gated Burn loader; keep the trait implementations out of vision_core to stay lean.

License

Apache-2.0 (see LICENSE in the repo root).