Pakistani Researchers Develop AI Technique to Determine Citrus Fruit Sweetness
-
Loading... - Jai Gandhi
- 14 Jun 2023
- 67 Views
- 0 Like
- 0 Comment
A group of Pakistani researchers has achieved a remarkable scientific breakthrough by creating a visual classification method based on artificial intelligence (AI) that accurately gauges the sweetness of indigenous citrus fruits.
Under the leadership of Dr. Ayesha Zeb from the National Centre of Robotics and Automation at the National University of Sciences and Technology (NUST), the team has successfully predicted fruit sweetness with an accuracy rate of over 80 percent, without causing any damage to the fruit during the process.
To carry out their experiment, the scientists carefully selected 92 citrus fruits, including varieties like Blood Red, Mosambi, and Succari, from a farm located in the Chakwal district. They employed a handheld spectrometer to capture spectra, which are patterns of light reflected from specific regions on the fruit's skin. Utilizing near-infrared (NIR) spectroscopy, a technique that enables the analysis of non-visible light spectra, the team examined the fruit samples. Out of the 92 fruits, 64 were used for calibration purposes, while the remaining 28 were used for prediction using the spectrometer.
Although the application of NIR spectroscopy in non-destructive fruit classification is not novel, the Pakistani team's approach involved its use in modeling the sweetness of local fruits. Moreover, they integrated artificial intelligence algorithms to directly classify the sweetness of oranges, resulting in enhanced accuracy.
Traditionally, determining fruit sweetness requires chemical and sensory testing. Oranges' sweetness is typically assessed by measuring total sugars, known as Brix, while levels of citric acid are indicated by titratable acidity (TA). To develop the AI model, the team obtained reference values for Brix, TA, and fruit sweetness by peeling off samples from the marked areas used for spectroscopy.
Laboratory testing of the juice extracted from the samples provided actual Brix and TA values. Additionally, human volunteers tasted the fruits and categorized them as flat, sweet, or very sweet.
Using the acquired spectrum, reference values, and sweetness labels, the team trained the AI algorithm using a total of 128 samples. The AI model was designed to predict Brix, TA, and sweetness levels based on spectral data. To assess the model's accuracy, the researchers tested it with data from 48 new fruits, comparing the predicted values with actual measurements obtained through sensory evaluations and chemical analysis.
The results were astounding, as the AI model not only accurately predicted the values of Brix, TA, and overall sweetness but also surpassed traditional methods in sweetness prediction. The model achieved an impressive overall accuracy rate of 81.03 percent in identifying sweet, mixed, and acidic tastes.
The project was a collaborative effort led by Dr. Ayesha Zeb and Dr. Mohsin Islam Tiwana from the National Centre of Robotics and Automation at the National University of Sciences and Technology (NUST). It also involved Dr. Waqar Shahid Qureshi from the School of Computer Science at Technological University Dublin, Ireland; Dr. Abdul Ghafoor, Dr. Muhammad Imran, and Dr. Alina Mirza from the Military College of Signals at NUST; Dr. Amanullah Malik from the Institute of Horticultural Sciences at the University of Agriculture, Faisalabad; and Dr. Eisa Alanazi from the Department of Computer Science at Umm Al-Qura University, Makkah, Saudi Arabia.
