Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

You can implement following pipeline:

1) Threshold scan image with Otsu's method (will works better if background will be darker than the blank);

2) Make morphological filtration to reduce the thresholding noise;

3) Fing good features by means of this goodfeaturestotrack;

4) Find minimal area rectangle for the good features;

5) Perform image rotation and cropping by means of found min area rect angle and sizes

You can implement following pipeline:

1) Threshold scan image with Otsu's method (will works better if background will be darker than the blank);

2) Make morphological filtration to reduce the thresholding noise;

3) Fing Find good features by means of this goodfeaturestotrack;

4) Find minimal area rectangle for the good features;

5) Perform image rotation and cropping by means of found min area rect angle and sizescropping