getImagePrediction static method
Get the inference result of a static image.
Using the image file image
(required) for inference, the image size for inference inputWidth
, inputHeight
,
the mean mean
and standard deviation std
for image normalization,
the threshold of the inference result minScore
, and get the inference result.
The format is List of
{ "rect": { "left": double, "top": double, "right": double, "bottom": double }, "mask": Uint8List, "keypoints": [[double, double], [double, double], [double, double], [double, double], ...], "confidenceInClass": double, "detectedClass": String }
.
"mask" and "keypoints" do not exist on some models.
Implementation
static Future<List> getImagePrediction({
required File image,
int inputWidth = kInputWidth,
int inputHeight = kInputHeight,
List<double> mean = kNormMean,
List<double> std = kNormStd,
double minScore = kMinScore,
}) async {
final List prediction = await _channel.invokeMethod(
'predictImage',
{
'image': image.readAsBytesSync(),
'inputWidth': inputWidth,
'inputHeight': inputHeight,
'mean': mean,
'std': std,
'minScore': minScore,
},
);
return prediction;
}