Add "fake" static capture method to allow both calibrator and OCR to run on sample frames
Add lots of easiercap sample frames for das trainer to be used as possible inputs
Lots of updates on calibrator:
add color preview, black and white preview, das previews
Add highlight for instant das
Use the split highlight for instant das and current piece das
Allow source id to be changed for static input, so sample frames can be quickly selected
refresh all images on tab changes (fixes behaviour of having missing images on tabs > 0)
Bunch of structural layout updates (add frame groups where logically relevant)
Remove the double scaling, capturing from input is as-is, and scaling is only to inject into the Image canvas
Add Das Trainer scanners
Note: the approach for the parseImage for das trainer's CurrentPiece is temporary, This PR is large-ish, so the earlier things can be merged the less chances of conflicts later. With the Current Piece Parser updatable later.
Context
NEStrisOCR is primarily for the default NES Tetris rom, but some useful mods outthere could be OCRed as well to be used to make great UIs.
This PR enabled Das Trainer to be OCR-ed
Approach
Note: the approach for the parseImage for das trainer's CurrentPiece is temporary, This PR is large-ish, so the earlier things can be merged the less chances of conflicts later. With the Current Piece Parser updatable later.
@blakegong @alex-ong