predict_image: Predict image
In surh/RosetteDetector: RosetteDetector

Takes an image file and a trained SVM model and predicts every pixel of the image.

1 2	predict_image(picture, m1, size_threshold = 0, predict_class = "plant", outmask = "outmask/", outoverlay = "outoverlay/")

`picture`	filename of image to be analyzed
`m1`	Trained SVM model from e1071
`size_threshold`	Minimum size in pixels that a feature must have to be kept
`predict_class`	Name of the class to be predicted. Must match a class name in m1.
`outmask`	Directory name where to save the masks resulting from the pipeline.
`outoverlay`	Directory name where to save the overlay images.