Using Smart Sensors to Measure Quality in Tomatoes and Mandarins Without Damage

Jenn Hoskins
27th August, 2024

Using Smart Sensors to Measure Quality in Tomatoes and Mandarins Without Damage

Image Source: Natural Science News, 2024

Key Findings

  • Researchers at Tanta University, Egypt, developed a non-destructive method to estimate fruit quality using image processing, spectral reflectance indices (SRIs), and machine learning
  • The study found that RGB indices and SRIs effectively estimated key quality parameters like chlorophyll, total soluble solids, and carotenoids in mandarins and tomatoes
  • Combining RGB indices and SRIs with machine learning models significantly improved the accuracy of fruit quality predictions, achieving high R2 values for both mandarins and tomatoes
Recent advancements in non-destructive techniques for assessing fruit quality have shown promising results, potentially revolutionizing the way we evaluate the ripeness and quality of fruits such as mandarins and tomatoes. Traditionally, fruit quality estimation has relied on destructive methods that are not only tedious and costly but also impractical for large-scale operations. A recent study by researchers at Tanta University, Egypt, aimed to address these challenges by utilizing image processing, spectral reflectance indices (SRIs), and machine learning models to estimate fruit quality parameters effectively[1]. The study focused on evaluating various quality parameters, including chlorophyll a (Chl a), chlorophyll b (Chl b), total soluble solids (TSS), titratable acidity (TA), the TSS/TA ratio, carotenoids (car), lycopene, and firmness. These parameters are critical indicators of fruit ripeness and overall quality. The researchers employed RGB indices and newly developed SRIs to quantify these parameters. The RGB indices are derived from the red, blue, and green color values of the fruit images, while SRIs are calculated based on the spectral reflectance of the fruits at different wavelengths. The findings demonstrated that both RGB indices and SRIs were highly effective in estimating various fruit properties. For instance, the relationship between RGB indices and measured parameters showed R2 values ranging from 0.62 to 0.96 for mandarins and from 0.29 to 0.90 for tomatoes. Specifically, the Visible Atmospheric Resistant Index (VARI) and Normalized Red (Rn) indices presented the highest R2 value of 0.96 with carotenoids in mandarin fruits. Similarly, the Excess Red Vegetation Index (ExR) showed an R2 value of 0.84 with carotenoids in tomato fruits. The study also explored the potential of combining RGB indices and SRIs with machine learning models such as Decision Tree (DT) and Random Forest (RF) to enhance the accuracy of fruit quality estimation. The DT-2HV model for mandarin fruits delivered exceptional results in predicting Chl a, with an R2 value of 0.993 and an RMSE (Root Mean Square Error) of 0.149 for the training set, and an R2 value of 0.991 with an RMSE of 0.114 for the validation set. For tomato fruits, the DT-5HV model achieved an R2 value of 0.905 and an RMSE of 0.077 for the training dataset, and an R2 value of 0.785 with an RMSE of 0.077 for the validation dataset. These results highlight the robustness of using RGB indices, SRIs, and machine learning models in non-destructive fruit quality estimation. The study's findings align with previous research that emphasized the importance of color and other external quality features in determining fruit maturity and quality. For example, a study on the plum variety 'Satluj Purple' found that color was the most dominant factor for classifying plums according to their maturity level, with strong associations between fruit acidity and mean intensity of green color (R2 = 0.9966)[2]. Furthermore, the research underscores the potential benefits of utilizing tomato byproducts, which contain bioactive compounds like lycopene that possess antioxidant, hypolipidemic, and anticarcinogenic activities[3]. By accurately estimating the quality parameters of tomatoes, this study could pave the way for more efficient use of tomato byproducts in functional foods, thereby addressing environmental concerns related to tomato waste disposal. In conclusion, the integration of image processing, spectral reflectance indices, and machine learning models offers a promising non-destructive approach to fruit quality estimation. The study conducted by Tanta University demonstrates that these techniques can provide accurate and reliable assessments of various quality parameters in mandarin and tomato fruits, potentially transforming the agricultural and food industries.

AgricultureBiotechBiochem

References

Main Study

1) Machine learning-driven assessment of biochemical qualities in tomato and mandarin using RGB and hyperspectral sensors as nondestructive technologies.

Published 26th August, 2024

https://doi.org/10.1371/journal.pone.0308826

Related Studies

2) Evaluation of plum fruit maturity by image processing techniques.

https://doi.org/10.1007/s13197-018-3220-0

3) Tomato and tomato byproducts. Human health benefits of lycopene and its application to meat products: a review.

https://doi.org/10.1080/10408398.2011.623799


Related Articles

An unhandled error has occurred. Reload 🗙