Solving the Inversion Problem with RGB Filters

1 reply · 2K views · last reply
Edward Zhang
Messages
1
Reaction score
0
I have some measurement data for known spectrum. Now I want to establish a algorithm for later measurement of unknown spectrum. I attached the equation in the image. r g b filters apply to the true spectrum I I am looking for. R G B are the measurement getting from the device. I1, I2, I3 are different spectrum measured. I want to based on the measured RGB values and known spectrum I1 I2 I3 to figure out r g b, so later I can use it for retrieve for other unknown I. I can figure out r g b with only one I, but I would thinking that's not accurate, so I want to bring in more known spectrum to form a training set. Anyone can help me with that? All the values are vectors contain 300 numbers.
[PLAIN][PLAIN][PLAIN]https://www.physicsforums.com/file:///C:\Users\zcj\AppData\Roaming\Tencent\Users\403532326\QQ\WinTemp\RichOle\_M4KG7`AMP48RLCUO%GG$WF.jpg
[PLAIN][PLAIN][PLAIN]https://www.physicsforums.com/file:///C:\Users\zcj\AppData\Roaming\Tencent\Users\403532326\QQ\WinTemp\RichOle\_M4KG7`AMP48RLCUO%GG$WF.jpg
[PLAIN][PLAIN][PLAIN]https://www.physicsforums.com/file:///C:\Users\zcj\AppData\Roaming\Tencent\Users\403532326\QQ\WinTemp\RichOle\_M4KG7`AMP48RLCUO%GG$WF.jpg
[PLAIN][PLAIN][PLAIN]https://www.physicsforums.com/file:///C:\Users\zcj\AppData\Roaming\Tencent\Users\403532326\QQ\WinTemp\RichOle\_M4KG7`AMP48RLCUO%GG$WF.jpg
[PLAIN][PLAIN][PLAIN]https://www.physicsforums.com/file:///C:\Users\zcj\AppData\Roaming\Tencent\Users\403532326\QQ\WinTemp\RichOle\_M4KG7`AMP48RLCUO%GG$WF.jpg
[PLAIN][PLAIN][PLAIN]https://www.physicsforums.com/file:///C:\Users\zcj\AppData\Roaming\Tencent\Users\403532326\QQ\WinTemp\RichOle\_M4KG7`AMP48RLCUO%GG$WF.jpg
 
Attachments
  • Matrix equation: a 3x3 matrix of measured RGB values R1 G1 B1 to R3 G3 B3 equals a 3x1 vector of spectra I1, I2, I3 times a row vector r g b.
    equation.jpg
    5.6 KB · Views: 565
Last edited by a moderator:
Thanks for the post! Sorry you aren't generating responses at the moment. Do you have any further information, come to any new conclusions or is it possible to reword the post?