Skip to content

README.md

Latest commit

 

History

History
179 lines (112 loc) · 3.16 KB

README.md

File metadata and controls

179 lines (112 loc) · 3.16 KB

PerchEye Flutter SDK – Dart API Reference

This class provides a high-level Dart interface for accessing native face recognition functionality through platform channels using the PerchEye Android SDK.

⚙️ Installation

Run this in your Flutter project:

flutter pub add perch_eye

🔌 Class: PerchEye

📡 Channel

static const MethodChannel _channel = MethodChannel('perch_eye_method_channel');

Internal method channel used for native communication (do not modify).

🧹 Methods

Future<void> init()

Initializes the underlying PerchEye native SDK. Must be called once before any recognition-related operations.

await PerchEye.init();

Future<void> destroy()

Releases SDK resources and unloads the model. Call this when the app or session ends.

await PerchEye.destroy();

Future<void> openTransaction()

Opens a new session for face processing. This is required before addImage().

await PerchEye.openTransaction();

Future<String> addImage(String base64)

Adds a face image to the current transaction for processing. Accepts a base64-encoded image string.

Returns:

  • A string representing the result enum (SUCCESS, FACE_NOT_FOUND, etc.)
final result = await PerchEye.addImage(base64Image);
if (result == 'SUCCESS') {
  // Proceed with enroll or verify
}

Future<String> enroll()

Generates a Base64-encoded face embedding hash from the previously added image.

Returns:

  • A String face hash
await PerchEye.openTransaction();
await PerchEye.addImage(base64Image);
final hash = await PerchEye.enroll();

Future<double> verify(String hash)

Verifies the last added image against a given face hash.

Returns:

  • Similarity score (0.0 to 1.0)
await PerchEye.openTransaction();
await PerchEye.addImage(base64Image);
final similarity = await PerchEye.verify(hash);

Future<String> evaluate(List<String> base64Images)

Evaluates multiple images to produce a single aggregated face hash.

Returns:

  • Combined hash as String
final hash = await PerchEye.evaluate([base64Image1, base64Image2]);

Future<double> compareList(List<String> images, String hash)

Compares a list of base64-encoded images against a face hash.

Returns:

  • Similarity score (0.0 to 1.0)
final sim = await PerchEye.compareList(
  [base64Image1, base64Image2],
  knownHash,
);

Future<double> compareFaces(String base64_1, String base64_2)

Convenience method: Internally performs:

  • openTransaction()
  • addImage(image1) → enroll() → openTransaction() → addImage(image2) → verify()

Returns:

  • Similarity score (0.0 to 1.0)
final similarity = await PerchEye.compareFaces(base64Image1, base64Image2);

🧪 Example Workflow

await PerchEye.init();

await PerchEye.openTransaction();
await PerchEye.addImage(base64Image1);
final hash = await PerchEye.enroll();

await PerchEye.openTransaction();
await PerchEye.addImage(base64Image2);
final similarity = await PerchEye.verify(hash);

await PerchEye.destroy();