AVSampleProcessor

public protocol AVSampleProcessor

To be implemented if custom image processing is needed. The default SwiftyTesseractRTE implementation does a small amount of image enhancement and conversion to grayscale. If the results from RealTimeEngine do not come back reliably, then performing your own image processing may be neccessary to receive optimal results.

Image transformation methods

  • Converts CMSampleBuffer into a grayscale UIImage.

    Declaration

    Swift

    func convertToGrayscaleUiImage(from sampleBuffer: CMSampleBuffer) -> UIImage?

    Parameters

    sampleBuffer

    The incoming CMSampleBuffer from the AVCaptureSession

    Return Value

    An optional grayscale UIImage

  • Crops UIImage to the bounds of areaOfInterest. The areaOfInterest must be located within the bounds of the AVCaptureVideoPreviewLayer or recognition will not be properly performed.

    Declaration

    Swift

    func crop(_ image: UIImage, toBoundsOf areaOfInterest: CGRect, containedIn previewLayer: AVCaptureVideoPreviewLayer) -> UIImage?

    Parameters

    image

    The image to be processed for OCR
    areaOfInterest

    The area within the AVCaptureVideoPreviewLayer to explicitly perform recognition on
    previewLayer

    Internal `RealTimeEngine

    Return Value

    Final UIImage ready for OCR