| Author | Message | ||
     Lars N�rgaard (lnofossdk) New member Username: lnofossdk Post Number: 1 Registered: 1-2011 |
Hi Joelle, Some of the issues you raise are dealt with in T�nning et al., Protein heterogeneity in wheat lots using single-seed NIT � A Theory of Sampling (TOS) breakdown of all sampling and analytical errors, ChemoLab, 84 (2006), 142-152. T�nning et al. Bulk Functionality Diversification by Unsupervised Single-Kernel Near-Infrared (SKNIR) Sorting of Wheat, Cereal Chemistry 86(6), (2009), 706-713. Pedersen et al., Near-infrared absorption and scattering separated by Extended Inverted Signal Correction (EISC). Analysis of NIT spectra of single wheat seeds. Applied Spectroscopy 56(9) (2002), 1206-1214. Nielsen et al., Development of non-destructive screening methods for single kernel characterization of wheat, Cereal Chem. 80(3), (2003), 274�280. You are welcome to contact me for a copy of these. Best regards Lars [email protected] | ||
| Joelle Rahm� (rahm�) New member Username: rahm� Post Number: 1 Registered: 10-2011 |
Dear all, I am new to NIR and I highly appreciate all the postings on this site. Your knowledge is a great help. I am working on single kernels of cereal grains and various beans. I use mainly reflectance in the range of 800-1600 nm. I was wondering if anyone here has experience with single kernel NIR and is willing to share it. Some of the challenges I face: - The effect of geometry of the grain - The use of light (e.g. When letting more light into the spectrometer, this should as far as I understand enhance signal to noise ratio. But isn�t there a risk to saturate and hide some information this way?) - Data treatment. One limitation is perhaps the need to average measurement replicates, since spectra from the same grain can highly differ. With a calibration that has been developed using several measurements per kernel, is it possible to make a prediction on a new kernel after measuring it only once? I appreciate all help and sharing. Thank you. Joelle |