Revision history [back]

What i tried is finding a mask by diferent ways and then overlapping them. For example try overlapping a mask found by thresholding hsv by value, other mask by Hue, other by saturación and then try adding them. It función to me.

Finding blobs also work, it found the object separately from is shadow.

Shadows are always gray, try a mask by Hue and then a mask by selecting a blob with more green like pixels.

And what i think is your best solution is using the same procedures microscope use, put the object betwen glasses and forget about shadows.

What Searching for the better way to do what you need y try inRange función after converting image to hsv, i tried is finding a mask by diferent ways use all hue and then overlapping them. For example try overlapping a mask found by thresholding hsv by value, other mask by Hue, other by saturación saturation range from 0 to 255 and then try adding them. It función value from 0 to me. 90.

Finding blobs also work, it found I work in Java but is almost the object separately from same. My result image is shadow.here:

Searching for the better way to do what you need y try inRange función after converting image to hsv, i use all hue and saturation range from 0 to 255 and value from 0 to 90.

I work in Java but is almost the same. My code is here https://github.com/greydelpf/opencv_functions/blob/master/hsv_bkg_sust.java and the result image is here:

Searching for the better way to do what you need y try inRange función after converting image to hsv, i use all hue and saturation range from 0 to 255 and value from 0 to 90.

I work in Java but is almost the same. My code is here https://github.com/greydelpf/opencv_functions/blob/master/hsv_bkg_sust.java and the result image is here: