Near-infrared Spectral Data Collected In The Calibration Study

Analytical applications of Near-Infrared spectroscopy technique what is rapid, nondestructive, environmental friendly and pollution-free have a significant role in the analysis of agricultural and food quality. As their own difficulties about Near-Infrared spectroscopy, our domestic products are not enough .Therefore, it is inevitable trend in researching the standards of Near-Infrared spectroscopy instruments that establishment and application of technology, and developing the basis of Near-Infrared on hardware and software.In recent years with the development of analytical applications of Near-Infrared spectroscopy technique, correction analysis is becoming more and more important as a basis of Near-Infrared spectroscopy technique. Correction analysis of this dissertation expands primarily in hardware and software, and combines with chemometrics software development to detect and verify the correction results.The correction to hardware is reflected in testing capabilities on samples quickly and accurately. Choosing the most suitable instrument models and the excellent performance are instrument correction range. The main task: summarize the main technical indicators on the instrument and how to select methods, then use physical test with Fourier transform spectrometer for drawing more accurate indicator.Software technology is improved mainly in the selection of wavelet transform and the selection method of modeling calibration. The threshold method is improved and applied to the near-infrared spectroscopy in the pretreatment, as well as the mixture PLS algorithm is used in the modeling calibration methods, and applied to practical applications in RIPP chemometrics software.In practical engineering, the improvement of pretreatment methods and the correction model algorithm are used in RIPP chemometrics software, and tested the two parts by chemometrics software. Then it successfully improves the analysis speed and accuracy, which verifies the accuracy of calibration.Rapid correction technology must be requested through the organic integration on the corrections to hardware and software technology. The study not only offers several suitable methods for the correction of the Near-Infrared spectroscopy, but also offers a strong support to improve the accuracy for Near-Infrared spectroscopy technique.