KARACHI: The creation of a visual categorization approach based on artificial intelligence (AI) that reliably determines the sweetness of local citrus fruits by a team of Pakistani researchers is a significant scientific advance.
The National University of Sciences and Technology (NUST) National Centre of Robotics and Automation team, under the direction of Dr. Ayesha Zeb, successfully forecasted fruit sweetness with an accuracy rate of over 80% without causing any harm to the fruit. The researchers chose 92 citrus fruits from a farm in the Chakwal area, including Blood Red, Mosambi, and Succari kinds, to use in their experiment. They used a portable spectrometer to collect spectra from defined areas on the fruits’ skin, which are patterns created by light bouncing off surfaces.
The researchers examined the fruit samples using near-infrared (NIR) spectroscopy, a method that permits the investigation of non-visible light spectra. 64 of the 92 fruits were utilised for calibration, and 28 were used for spectrometer-based prediction. NIR spectroscopy has been used to classify damage-free fruits for some time, but the Pakistani team used an innovative method by using it to simulate the sweetness of local fruits. The accuracy was increased by incorporating artificial intelligence algorithms for the direct categorization of orange sweetness. With 0.46 million tonnes of citrus fruit exports in 2020, Pakistan, the sixth-largest producer in the world, would benefit from this development.
- Advertisement -
- Advertisement -
- Advertisement -
- Advertisement -