cat_detection library

On-device cat detection and landmark estimation using TensorFlow Lite.

This library provides a Flutter plugin for cat detection using a unified multi-stage TFLite pipeline: SSD body detection, species classification, body pose estimation, and face landmark extraction.

Quick Start:

import 'package:cat_detection/cat_detection.dart';

final detector = CatDetector(mode: CatDetectionMode.full);
await detector.initialize();

final cats = await detector.detect(imageBytes);
for (final cat in cats) {
  print('Cat at ${cat.boundingBox} score=${cat.score}');
  if (cat.pose != null) {
    final tail = cat.pose!.getLandmark(AnimalPoseLandmarkType.tailEnd);
    print('Tail: (${tail?.x}, ${tail?.y})');
  }
  if (cat.face != null && cat.face!.hasLandmarks) {
    final nose = cat.face!.getLandmark(CatLandmarkType.noseTipLeft);
    print('Nose tip: (${nose?.x}, ${nose?.y})');
  }
}

await detector.dispose();

Main Classes:

  • CatDetectorIsolate: Background isolate wrapper for cat detection
  • CatDetector: Main API for cat detection
  • Cat: Top-level detection result with body, pose and face info
  • CatFace: Detected cat face with bounding box and landmarks
  • CatLandmark: Single face keypoint with 2D coordinates
  • BoundingBox: Axis-aligned rectangle in pixel coordinates

Detection Modes:

Pose Model Variants:

Face Landmark Model Variants:

Skeleton Connections:

Classes

Animal
Top-level result for a single detected animal.
AnimalDetector
On-device animal detection using a multi-stage TensorFlow Lite pipeline.
AnimalPose
Full-body pose result for a single detected animal.
AnimalPoseLandmark
A single body pose keypoint with 2D coordinates and a confidence score.
BoundingBox
An axis-aligned or rotated bounding box defined by four corner points.
Cat
Top-level result for a single detected cat.
CatDetectionDart
CatDetector
On-device cat detection using a unified multi-stage TensorFlow Lite pipeline.
CatDetectorIsolate
A wrapper that runs the entire cat detection pipeline in a background isolate.
CatFace
Detected cat face with bounding box and optional landmarks.
CatLandmark
A single cat face keypoint with 2D coordinates.
CropMetadata
Metadata for mapping landmark coordinates from crop space back to original image space.
Mat
ModelDownloader
Downloads and caches the HRNet body pose model from GitHub Releases.
PerformanceConfig
Configuration for interpreter hardware acceleration and threading.
Point
A point with x, y, and optional z coordinates.

Enums

AnimalPoseLandmarkType
SuperAnimal body keypoint types (indices 15–38 in the full SuperAnimal topology).
AnimalPoseModel
Body pose model variant for SuperAnimal keypoint extraction.
CatDetectionMode
Detection mode controlling the full pipeline behavior.
CatLandmarkModel
Cat landmark model variant for landmark extraction.
CatLandmarkType
Cat face landmark types based on the CatFLW dataset topology.
PerformanceMode
Hardware acceleration mode for LiteRT inference.

Constants

animalPoseConnections → const List<List<AnimalPoseLandmarkType>>
Defines the standard skeleton connections between SuperAnimal body keypoints.
catLandmarkConnections → const List<List<CatLandmarkType>>
Defines the standard skeleton connections between cat face landmarks.
catLandmarkFlipIndex → const List<int>
Landmark index permutation for horizontal flip (CatFLW convention).
IMREAD_COLOR → const int
numCatLandmarks → const int
Number of cat face landmarks (48 for the CatFLW model).

Functions

imdecode(Uint8List buf, int flags, {Mat? dst}) → Mat
imdecode reads an image from a buffer in memory. The function imdecode reads an image from the specified buffer in memory. If the buffer is too short or contains invalid data, the function returns an empty matrix. @param buf Input array or vector of bytes. @param flags The same flags as in cv::imread, see cv::ImreadModes.