Near-Infrared Spectroscopy For Sugar Content Evolution In Potatoes Over Multiple Growing Seasons

INTRODUCTION

Potato is an essential crop that contributes nutritional value including carbohydrates, protein, and vitamins to the human diet in both developed and developing countries. A famous solution was applied in 1980 using NIR spectroscopy to determine the hygienic content in potatoes. In this application note, we focus on this method for monitoring sugar concentration in potatoes.

Plenty of publications describe achievements in various industries to ensure the sugar content in commercial potato products provides recommended sugar levels. Crisp industries utilize NIR spectroscopy to supervise consistent cooking time by preserving chip sugar value. This method also expresses the same use in dried potato and quality-checking industries.

The analysis classifies models relying on glucose and sucrose concentrations for potato tubers (of Frito Lay 1879 (FL), a chipping cultivar, and Russet Norkotah (RN). The classification is performed to identify tubers with either high or low concentrations. A combination of optical measurements and machine-learning models will assist in the definition of sugar content rapidly and effectively, with non-destructive features extremely preferred in manufacturing processing. The testing models are built up with data from different seasons of harvesting condensing accuracy for the prediction.  

METHODOLOGY

The methodology was distinguished into two functional sections: optical experiments and data processing, which will shed light on the paragraphs below.

Diffuse reflectance spectra were investigated in a range from 900 to 1700 nm. Herein, the investigation concentrated on the vision of absorbance signal was defined by log(1/reflectance). Determination of potato in absorbance spectra should contain characteristic peaks of NIR spectral region primarily related to overtones and combinations of fundamental vibrations of C-H, N-H, and O-H groups. The absorption peaks worth 970, 1200, and 1450 nm (Figure 1a) were noted in the whole tubers in both cultivars. To clarify the focusing analyzed peaks, 970 nm is associated with O-H overtones, indicative of water content, 1200 nm is related to C-H and O-H combination bands, and 1450 nm is another O-H band, significant for water content determination. The results would show the same trend as in different seasons.

After attaining the database, the technician would process data suitable for each machine learning type, then the prediction would be applied. The fitting line between the calibrated and measured lines could match over 99%, which means the predicted fulfillment could be up to 99% accurate, even with FL or NR.

Figure 1b describes the results of cross-validation carried out to evaluate the robustness of the obtained models in an actual test. Internal validation demonstrated superior results for dry matter, with higher R-squared values and favorable prediction errors, compared to reducing sugars. The SEP, SEP(C), and bias metrics confirm that the calibration models for both quality parameters are effective. Thus, the results highlight the models' robust prediction capabilities and reliability.

Figure 1. a) NIR spectra were measured by potatoes; b) Comparison of reference values with predicted values by the NIRS model for each quality parameter. RSQ: Multiple correlation coefficient; SEP: Standard error of prediction; SEP (C): Standard error of prediction corrected by BIAS.

RECOMMENDED EQUIPMENT

The concept of the NIR spectroscope is familiar to consumers and easy to equip quickly by contacting Intins, we could provide the customers with a complete system for measuring reflection. Intins would recommend the customers use the products Ocean Insight, USA manufactured. A NIRQuest spectrometer and a Vis-NIR light source are the best choice for this application.

The very first appearance of this system is the spectrometer, which plays a crucial role in the precision of the measurement with outstanding characteristics as well as pre-configured models across 900-2450 nm with an option of 900-1650 nm that serves this application requirement. Its optical bench design for higher-sensitivity performance is well-suited for your NIR application. Especially, this special spectrometer also accepts moisture content for fruit in general and potatoes in particular. In addition, the NIRQuest contains a marvelous InGaAs array detector, which expresses high performance acquiring data as fast as 5 milliseconds, and also temperature-regulated being internally cooled for optimum signal-to-noise and sensitivity. The data is returned 64 bytes (USB 1.1) or 512 bytes (USB 2.0). The trigger mode was integrated with this product line.

Intins provides robust and reliable tungsten halogen light sources tailored to meet your application needs. The HL-2000 product family offers a range of options, from high-power to long-lifetime models, ensuring flexibility for laboratory measurements across the 360-2400 nm spectrum. All HL-2000 models are equipped with an integrated fan to maintain a cool and stable light source, and feature a built-in holder for accommodating filters to condition the light. Some models also include an integrated shutter and long-lifetime bulb. Additionally, a universal power supply simplifies and expedites the setup process.

CONCLUSION

NIR spectroscopy offers a powerful, non-destructive solution for monitoring sugar concentration in potatoes, essential for maintaining quality in crisp and dried potato products. Utilizing Ocean Insight's NIRQuest spectrometer and accessories, this method ensures high accuracy across harvest seasons, achieving over 99% prediction accuracy. This reliable technology supports consistent product quality and efficient processing, making it an indispensable tool for the potato industry.

The very first appearance of this system is the spectrometer, which plays a crucial role in the precision of the measurement with outstanding characteristics as well as pre-configured models across 900-2450 nm with an option of 900-1650 nm that serves this application requirement. Its optical bench design for higher-sensitivity performance is well-suited for your NIR application. Especially, this special spectrometer also accepts moisture content for fruit in general and potatoes in particular. In addition, the NIRQuest contains a marvelous InGaAs array detector, which expresses high performance acquiring data as fast as 5 milliseconds, and also temperature-regulated being internally cooled for optimum signal-to-noise and sensitivity. The data is returned 64 bytes (USB 1.1) or 512 bytes (USB 2.0). The trigger mode was integrated with this product line.