Iran has developed a smart screening system capable of detecting and separating pistachios suspected of aflatoxin contamination at a speed of 100 frames per second,
IRNA, a partner of TV BRICS, reported.
The system was designed by a knowledge-based company based at the Science and Technology Park of the University of Tehran and combines robotics, machine vision and artificial intelligence technologies for applications in food and agriculture.
Mohammad Reza Farahnak, the company’s chief executive, said the system was developed in response to quality-control needs in the pistachio and dried-fruit sector. Its development is based on deep learning and simultaneous imaging in visible and ultraviolet light.
The system combines BGY fluorescence in the UV spectrum, imaging at a rate of 100 frames per second and deep-learning algorithms. It can detect small-scale defects measuring 0.1 millimetres on production lines and use a pneumatic system to separate suspected contaminated products from the main flow.
He said the system’s development required extensive data collection because of the diversity and complexity of visible and fluorescent signs associated with aflatoxin contamination.
“The diversity and complexity of visible and fluorescent signs required extensive data collection and close cooperation with pistachio industry stakeholders,” Farahnak said, noting that the technical team subsequently developed accurate and reliable models.
The system integrates mechanical design, electronics, control systems and artificial intelligence. Farahnak said this approach allows the equipment to be optimised according to specific requirements and real production-line conditions. Remote monitoring and control capabilities can also reduce equipment downtime and facilitate industrial operation.
Looking ahead, the team plans to enter export markets and develop similar systems for other agricultural products, food products, recycling, and industrial and mineral quality control.