0

我firebase_ml_vision用来识别人脸。我需要从图像中提取人脸。边界条件给出了firebase人脸检测方法。

我使用来自https://pub.devcopyCrop的方法Image插件来提取人脸。我能够保存结果。但是当我尝试打开图像时，我得到了这些错误。我尝试保存两种格式和.PNGJPEG

JPEG 格式

PNG格式

这是我写的函数。


import 'package:image/image.dart' as editImage;

_editImage() async {
    final Face face = await _scanResults[0];
    final left = face.boundingBox.left.truncate();
    final right = face.boundingBox.right.truncate();
    final top = face.boundingBox.top.truncate();
    final bottom = face.boundingBox.bottom.truncate();

    print("Boundings" + face.boundingBox.toString());

    Directory d = await getExternalStorageDirectory();
    DateTime date = DateTime.now();

    File img = File(_path);
    editImage.Image originalImage =
        editImage.decodeImage(img.readAsBytesSync());
    editImage.Image faceCrop =
        editImage.copyCrop(originalImage, left, top, right, bottom);
    final encodedImg = editImage.encodeJpg(faceCrop);

    final decodedImg = await decodeImageFromList(encodedImg);
    final result = await decodedImg.toByteData();

    final finalPath = "${d.path}/Face_Image_${date.toIso8601String()}.png";
    final finalImage = await new File(finalPath).create();
    final finalResult =
        result.buffer.asUint8List(result.offsetInBytes, result.lengthInBytes);
    finalImage.writeAsBytesSync(finalResult);
    print("Face crop finished");

  }
4

0 回答 0