Processing screen captures

CSDK contains special preprocessing algorithm that is optimized for processing screen captures. This means that different binarization and resolution enhancement is applied and neither deskew nor autorotation is applied. This preprocessing algorithm works only on color and grayscale images.

In order to process screen captures it is important that the original image to be without loss, so the image must be passed to CSDK in uncompressed or lossless image format. Therefore the input image file cannot be JPEG, TIFF-JPG and GIF (image handling applications may use dithering algorithm when the image is saved to GIF file). We suggest to use PNG or TIF LZW.

If the images to be processed can include screen capture, you need to set Kernel.Img.ScreenCapture to SCR_AUTO or SCR_YES. We suggest to use SCR_AUTO, because in this case CSDK analyzes the image during preprocessing, and decides whether the normal or the screenshot-optimized algorithm will be applied.