Your shopping cart is empty!
Fruit Online Sorting Based on NIR Spectroscopy
With the improvement of the quality of life and the change of consumption level, the different demands of consumers for fruit quality have also contributed to the classification of fruit sales. However, traditional destructive inspection methods are not only costly, but also cause waste of resources. Therefore, the method of spectral non-destructive inspection has become a major trend. So, why use NIR spectroscopy to perform fruit sorting?
When irradiating fruits with near-infrared light, the internal components of different fruits have different degrees of optical absorption and scattering for different wavelengths, and the internal spectrum will also change with the mass fraction of the internal components of the fruit. Using this feature, the main components and their mass fractions in fruits can be analyzed based on the characteristics of near-infrared spectroscopy.
Webpetsupply recommends ATP8600 which is low cost NIR spectrometer, combined with Tungsten Halogen lamp ATG1002, Raman Probe holder-X,Y,Z axizs and Webpetsupply free software, form a complete fruit online sorting and detection system.
Detect fruit sugar content based on near-infrared spectroscopy technology (Moisture/Black-heart Disease [visible + near-infrared])
The main process:
(1) Select representative fruits
(2) Collect relevant spectral data of fruit samples by diffuse reflection or transmission;
(3) Preprocess the spectral data, eliminate errors caused by different factors on the accuracy of the fruit model, and select spectral data of more representative samples;
(4) Use national and internationally certified chemical analysis methods to measure the accurate content of fruit samples;
(5) Establish a prediction model
(6) Collect near-infrared spectra of unknown fruit samples, and then use the established prediction model to predict the component content of unknown samples.
(7) Use standard chemical analysis methods to measure the content of unknown fruit samples, verify the accuracy of the established prediction model, and then correct and optimize the prediction model.
Search
Categories
Popular Posts
Latest Posts
Comments: 0
No comments