OCR can be done to all images. This may or may not happen at the same time as document layout analysis.
If baseline detection is done, for circular objects (e.g. CDs), rotation can be applied to the image, to improve the legibility and ultimately the embedding of detected segments.
OCR can be done to all images. This may or may not happen at the same time as document layout analysis.
If baseline detection is done, for circular objects (e.g. CDs), rotation can be applied to the image, to improve the legibility and ultimately the embedding of detected segments.