You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hey,
This is a challenging question due to multiple factors:
The rectangles may have different rotations in images,
The colors may have variations and near-black colors may exist in images,
The background is also black which makes the process more difficult.
Thus, in order to be able to get the mask of bars and decode them, you should decide whether you want to do it automatically or human interaction is acceptable. If you can use user inputs, then, you can ask for a bounding box around the bars, and after than, perform a simple thresholding operation. Otherwise, in case of automatic operation, you may have to do some more steps. I think, some morphological closing operations combined with a differencing operation could result in acceptable outputs.
The text was updated successfully, but these errors were encountered: