Anzeige

How Sweet It Is!

Brix Estimation of Oranges using Hyperspectral Imaging (HSI)

An installation of a Headwall HSI system running perClass Mira spectral analysis software at a major orange juicing factory in Florida demonstrates the potential of HSI.

Bild: Headwall BVBABild: Headwall BVBA

123 Sweetness, or in the case of fruit juices, the concentration of soluble solids that are mostly sugars is of particular importance. Commonly reported in degrees Brix, sweetness is a key quality factor in assessing the grade of the fruits delivered to the processing facility. Results are used to calculate the value of the delivered fruit and are directly tied to costs and profitability. The measurement traditionally requires obtaining a representative sample from the delivered load and then juicing and analyzing the sample using a laboratory refractometer. However, the process is slow and labor intensive. Hyperspectral Imaging (HSI) offers a solution to this challenge. An HSI system can rapidly collect spectral data from a load of oranges passing under the sensor and enable calculation of the Brix values for a truckload or even individual oranges. Recently a Headwall Hyperspec MV.X VNIR (400-1000nm) HSI system running perClass Mira spectral analysis software was installed at a major orange juicing factory in Florida and demonstrates the potential of HSI. Incoming truckloads of oranges are scanned as they pass under the HSI system. The edge-processing MV.X system analyzes the data and delivers a predicted °Brix value for each truckload in real time to production control stations. Data processing is performed in two steps: First, a classification algorithm is used to detect and separate any foreign material interspersed with the oranges. Only HSI data coming from identified oranges is then forwarded to a regression model calculating an average °Brix value for the entire truck load. The regression model is based on reference values generated from juiced oranges for which °Brix was measured using a digital refractometer. The refractometer data was used to train a regression model to predict °Brix based on hyperspectral data collected from the oranges before juicing. A plot of the measured versus calculated °Brix demonstrates a good level of agreement between the methods with an R2=0.98. The resulting regression model was then uploaded for real-time execution on the MV.X HSI system. The installation of the fully automated HSI system has significantly improved the efficiency of inspection of the incoming oranges that was previously performed manually for a much smaller sample set.

Headwall BVBA

Dieser Artikel erschien in inVISION 1 (März) 2022 - 08.03.22.
Für weitere Artikel besuchen Sie www.invision-news.de