Open jonastheman opened 5 years ago
I use the training app with font 'Verdana', let it run for 2 minutes minimum. Save the OCR-Network file and use it to replace it the one in my real app('/Pods/SwiftOCR/framework/SwiftOCR/OCR-Network').
I take a photo in the app and runs:
swiftOCRInstance.recognize(myImage) { recognizedString in DispatchQueue.main.async { self.textResult.text = recognizedString } <<<<
Fails badly with translating simple image with Verdana font text on it.... What am I doing wrong?
Try zooming in on the image before running ocr - minimal crop around text part of image.
I use the training app with font 'Verdana', let it run for 2 minutes minimum. Save the OCR-Network file and use it to replace it the one in my real app('/Pods/SwiftOCR/framework/SwiftOCR/OCR-Network').
I take a photo in the app and runs:
Fails badly with translating simple image with Verdana font text on it.... What am I doing wrong?