DSP Robotics Support

I am working on a project similar to RSpec (http://www.rspec-astro.com/) and VSpec (http://www.astrosurf.com/vdesnoux/) which are used to evaluate amateur astro spectrum captured with a spectrometer and a telescope. I have mocked up a concept using the Image scanner schematic (unsure of author), Nubeat's scroll/zoom schematic, and the RGB to HSL schematic (mwvdlee, I think). It is suppose to work like this:

1) Import a spectrum.
2) Match to a reference spectrum by aligning with the scroll/zoom feature
3) Scan the spectrum to extract color luminescence, RGB values and plot against spectrum wavelength
4) Compare plot to determine parameters such as temperature and possible atomic signatures

However, I am having trouble understanding the image scanner portion of the schematic. Ideally, we need the pixel color values across the entire spectrum image (length of image). However, the schematic doesn't appear to do that and may limit to only 100 data points. I have tried various changes but they also do not appear to provide the desired data.

The graph/plot part of the schematic are just to see how values and things change. They still need a lot of work and do not accurately represent the spectrum.

Would anyone be able to provide some guidance on the image scanner part of the schematic?

It may be easier to rebuild the scanner in Ruby rather than trying to decipher the green spaghetti code.

martinvicanek - I have been resistant to study and utilize ruby for some reason and instead have been focusing on learning C++.

I think I may have made some progress. After studying the color scanner module and placing value tracking at various points, I was able to get it to scan the pixels of the entire width of the spectrum image and the graphs make more sense now. The RGB graph in the lower left shows where the max peak of each of the colors would be expected to fall and plots the actual emission lines as individual peaks. The HSL conversion plots just the luminescence value and therefore shows only the emission lines (graph upper right and lower right).

Now I just need to figure out how to calibrate the spectrum so proper wavelengths can be assigned. Then using black body radiation equations, temperatures could be calculated.

Still needs a lot of work.