Ah, in that case a multi-band sensor like the one you located might work pretty well. But you will likely need to calibrate it with actual test strips. Your color chart reference will have been made with half-toned inks to give the same visual appearance as the actual test strip, but there will likely be significant spectral differences that a multichannel sensor will see differently. The manufacturer of the test strips might have actual spectral response data on the dyes in the test strip that could also be the basis for working out a proper calibration process. if your project has significant backing then the purchase of a real spectral-photometer might be in order, that would allow you to get the same information directly.
You will also need to pay close attention to the spectra of whatever light source is used to illuminate the sample. Traditionally incandescent bulbs are nearly always used for this purpose because of their uniform spectra. It might be possible to use a white LED light source for your project, but the very uneven spectra is going to complicate things.