Note that this visualization is very handy when debugging. You can do similar debugging techniques when working with normal vectors, where each normal vector (3 dimensions) can be mapped to red (R), green (G), and blue (B). It's a nice way to get a feel for correspondences with intuitive coloring.
Zc0in
Can we use HSV or lab* instead of RGB? I learned that lab* is intended as a perceptually uniform space, which is a better representation of color space and it is designed to approximate human vision. Why we still prefer rgb but not lab*? Is it not handy enough?
sZwX74
@Zc0in From what I understand, the RGB model is more used for computers because monitors produce color by emitting the appropriate amount of R, G, and B light per pixel, and since that's how monitors were originally designed, that's become the default for how computers process color. In practice, all the different color schemes are used (I know HSV is used in CV because traveling along each of its axes provides more valuable information traveling along the RGB axes)
austinapatel
How does one determine the mapping between the 3D space and the 2D texture space? Since the texture is distorted when mapping from 2D to 3D, is there some sort of "optimal" layout that can be used when laying out the 2D texture map to minimize distortion?
Note that this visualization is very handy when debugging. You can do similar debugging techniques when working with normal vectors, where each normal vector (3 dimensions) can be mapped to red (R), green (G), and blue (B). It's a nice way to get a feel for correspondences with intuitive coloring.
Can we use HSV or lab* instead of RGB? I learned that lab* is intended as a perceptually uniform space, which is a better representation of color space and it is designed to approximate human vision. Why we still prefer rgb but not lab*? Is it not handy enough?
@Zc0in From what I understand, the RGB model is more used for computers because monitors produce color by emitting the appropriate amount of R, G, and B light per pixel, and since that's how monitors were originally designed, that's become the default for how computers process color. In practice, all the different color schemes are used (I know HSV is used in CV because traveling along each of its axes provides more valuable information traveling along the RGB axes)
How does one determine the mapping between the 3D space and the 2D texture space? Since the texture is distorted when mapping from 2D to 3D, is there some sort of "optimal" layout that can be used when laying out the 2D texture map to minimize distortion?