Hi there,
I was wondering if there's any open software (besides OpenRaman's) that will convert a CCD image into a Raman spectrum. If not, what technique is used to do this so that I can just code it myself.
Thank you,
Bowen
Converting CCD Image to Raman Spectrum
Re: Converting CCD Image to Raman Spectrum
I don't have any names on the top of my head but there are surely some. At the very least, you can use the Matlab code that I gave here : http://www.open-raman.org/build/breadboard/assembly/
The process of converting an image to a spectra is not very complicated and can be split into two steps:
(1) Conversion of the image to spectral data. In this step, you just sum the columns of the image to yield an output vector. You can add pre-processing and post-processing filtering steps to that to clean up the spectra as well. Typically, I perform a median filtering on the image before integrating the columns and I do a box filter on the resulting vector data.
(2) Calibration of the x-axis of the spectra. Up to now, you should have a relatively clean spectral vector but you don't know what wavelength correspond to which entry. To perform this conversion, the best is to acquire a spectra of a known source where you can assign peak to actual wavelength values. I can recommend reading this post to understand how it's done : https://www.thepulsar.be/article/robust ... trometers/
And that's all! No magic involved
The process of converting an image to a spectra is not very complicated and can be split into two steps:
(1) Conversion of the image to spectral data. In this step, you just sum the columns of the image to yield an output vector. You can add pre-processing and post-processing filtering steps to that to clean up the spectra as well. Typically, I perform a median filtering on the image before integrating the columns and I do a box filter on the resulting vector data.
(2) Calibration of the x-axis of the spectra. Up to now, you should have a relatively clean spectral vector but you don't know what wavelength correspond to which entry. To perform this conversion, the best is to acquire a spectra of a known source where you can assign peak to actual wavelength values. I can recommend reading this post to understand how it's done : https://www.thepulsar.be/article/robust ... trometers/
And that's all! No magic involved