New answers tagged machine-learning
0
My answer focuses on the machine learning part instead of the original question concerning LAB color space.
From a Machine Learning perspective (and more specifically Deep Learning with semantic segmentation networks for your application) taking as input all of the four bands seems optimal unless you can assert that one (or more) channel(s) is (are) not ...
1
Let's start with the source of your RGB + near IR images. To break the process down:
Some camera must have taken those images.
The camera sensor is sensitive to light within two given wavelengths. The technical term for this is spectral response.
Alternating elements (Bayer pattern) in a sensor are etched with filters. Each filter type passes certain ...
Top 50 recent answers are included
