It would be great if we could extend an image control, so that we have the ability to manually crop an image by manually defining the boundary of the crop. Then to save the output (i.e. the cropped image).
Where I work, we built a custom canvas app in PowerApps to scan and track our assets. However, sometimes it isn't possible and so we are looking at using the OCR capabilities to extract serial numbers. However, OCR is going to extract all text inside the image. While I can use certain functions to help identify the serial number, giving the user the ability to focus their image on the serial number to eliminate unwanted characters would be a more elegant solution, IMO.
Please make this idea an option.
A "manually cropped" image would be ideal, but if not possible, or maybe an "small step ahead" would be taking picture using a mask, like the one exist on any "barcode scanning" apps, so an user could "fit" the required part in that mask.