PAK Scientists Develop AI Method to Determine Citrus Fruit Sweetness:
A breakthrough has been achieved by a team of Pakistani scientists who have developed an artificial intelligence (AI) method to accurately determine the sweetness of local citrus fruits without causing any damage. Led by Dr. Ayesha Zeb from the National Centre of Robotics and Automation at the National University of Sciences and Technology (NUST), the team achieved an impressive accuracy rate of over 80% in predicting fruit sweetness. To conduct their experiment, the researchers carefully selected 92 citrus fruits, including varieties such as Blood Red, Mosambi, and Succari, from a farm located in the Chakwal district.
Methodology:
In their methodology, the team utilized a handheld spectrometer to collect patterns of light reflecting off marked regions on the fruit’s skin, employing a technique known as spectra. Near-infrared (NIR) spectroscopy, which analyzes non-visible light, employed to examine the fruit samples. Out of the 92 fruits, 64 were use for calibration purposes, while the remaining 28 were use for prediction, utilizing the spectrometer. While the use of NIR spectroscopy for non-destructive fruit classification is not novel, the unique aspect of the Pakistani team’s approach lies in its application for modeling the sweetness of local fruits. They also incorporated artificial intelligence algorithms to directly classify the sweetness of oranges, resulting in improved accuracy.
Traditionally, assessing fruit sweetness involves chemical and sensory testing. In the case of oranges, sweetness determined by measuring the total sugars, referred to as Brix, while citric acid levels are indicate by titratable acidity (TA). To develop their 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 tests conducted on the extracted juice from the samples to obtain the actual Brix and TA values. Additionally, human volunteers tasted the fruits and categorized them as flat, sweet, or very sweet.
Using the collected spectra, reference values, and sweetness labels, the team trained the AI algorithm utilizing a total of 128 samples. The AI model designed to predict Brix, TA, and sweetness levels based on the spectral data. To assess the accuracy of the model, the researchers tested it with data from 48 new fruits, comparing the predicted values with actual measurements obtained through sensory evaluations and chemical analysis.
You may like our posts:
“Pakistan Army is One of the Strongest Forces in the World”: Army Chief Gen Asim Munir
Roosevelt Hotel of PIA Leased Out to New York City Firm for 3 Years
Govt Employees Using Free Electricity Worth Around Rs. 8 Billion in Pakistan
Here’s Why Afghanistan Currency is Stronger than Pakistan’s Currency
