Package com.codename1.ai.vision

package com.codename1.ai.vision

Vendor-neutral on-device vision APIs for still images and live camera frames.

The automatic backend uses Apple Vision/Core Image on iOS and Mac Catalyst and ML Kit on Android. Unsupported ports report that through isSupported(). Optional backends are selected with VisionBackends.

Each analyzer is a separate build-time feature. Referencing one analyzer causes the builder to retain only its platform adapter and native dependency. VisionPipeline safely copies callback-owned camera frames and drops stale pending frames under load.

Related Packages

Package	Description
com.codename1.ai	Codename One's AI / LLM client surface plus the value types, streaming primitives, and chat-binding helpers that sit on top of it.
com.codename1.ai.inference	Reusable on-device inference sessions for .tflite models.
com.codename1.ai.language	Vendor-neutral on-device language identification, translation, and smart reply.

Class	Description
Barcode	Portable barcode observation with normalized geometry.
BarcodeScanner	Creates reusable still-image or live-frame barcode analyzers.
DocumentScanner	Creates reusable document-boundary and perspective-correction analyzers.
DocumentScanResult	Corrected document pages returned as encoded image data.
Face	Portable face observation.
FaceDetector	Creates reusable on-device face analyzers.
ImageLabel	Portable ranked image-classification label.
ImageLabeler	Creates reusable on-device image classifiers.
Pose	Portable body-pose result.
Pose.Landmark	One named body joint with normalized position and confidence.
PoseDetector	Creates reusable body-pose analyzers.
SegmentationMask	Dense per-pixel foreground confidence mask.
SelfieSegmenter	Creates reusable foreground/person segmentation analyzers.
TextRecognitionResult	OCR output containing the complete recognized text and portable block-level geometry.
TextRecognitionResult.TextBlock	One recognized text block with confidence and normalized bounds.
TextRecognizer	Creates reusable on-device OCR analyzers.
VisionAnalyzer<T>	Reusable, closable on-device analyzer for still images or camera frames.
VisionBackend	Identifies a vision implementation.
VisionBackends	Vision backend selectors.
VisionException	Failure reported by an on-device vision backend.
VisionFeature	Features understood by the shared vision backend.
VisionImage	Immutable input for encoded still images or raw camera pixels.
VisionMetadata	Optional backend identity and backend-specific diagnostic strings attached to a vision result.
VisionOptions	Common analyzer configuration.
VisionPipeline<T>	Connects a camera frame stream to a reusable analyzer with keep-only-latest backpressure.
VisionPipelineListener<T>	Receives live vision results and recoverable analysis failures on the EDT.
VisionPoint	Immutable point in a normalized, top-left-origin coordinate space.
VisionRect	Immutable normalized rectangle using a top-left origin.