Artificial-intelligence-driven nondestructive detection and Internet of Things monitoring for improving the postharvest quality of fruits and vegetables

Introduction: Artificial intelligence is a powerful tool that can be used to support the sensing detection and monitoring of the postharvest quality of fresh fruits and vegetables, which can ensure product standardization and high value, reduce postharvest losses, and improve the comprehensive utilization rate.

Methods: The optical transmission characteristic parameters of fruit and vegetable tissue were calculated using the reverse doubling method, and the light scattering characteristics, penetration depth, and energy field distribution of fruit and vegetable tissue were analyzed. The correlation between optical characteristic parameters and fruit and vegetable quality was established using machine learning, which provided method support for the structural design of a fruit and vegetable quality nondestructive detection system. On the other hand, a fruit and vegetable storage monitoring and early warning system integrating temperature, humidity, and gas sensors was developed, and a dynamic monitoring and early warning model of the fruit and vegetable deterioration process was established using combining deep learning.

Results: The intelligent optimization algorithm was used to extract the spectral feature signal and was combined with the artificial intelligence algorithm. Moreover, the stable and high-precision prediction model of the key indicators of fruit and vegetable quality was established, and the correlation coefficient of the prediction models was found to be above 0.90. The spatio-temporal dynamics model (used to find the relationship between multi-source environmental factors and quality) and the early warning and discrimination model (used to identify the interaction between environmental factors) were established, and the early warning accuracy was found to be greater than 92%.

Conclusions: A series of fast, high-precision, nondestructive, online, and intelligent detection systems used for improving main quality and safety indexes of fruits and vegetables were developed. The offline detection time was less than 2 seconds and the online detection speed was 3 samples per second. IoT sensing and Internet technologies were established as intelligent real-time monitoring, evaluation, and early-warning technologies for fruits and vegetables. Artificial intelligence drives the intelligent, green, sustainable, and high-quality development of agricultural products.