Towards the Integration of Computer Vision and Applied Artificial Intelligence in Postharvest Storage Systems: Non-invasive Harvested Crop Monitoring

Abstract

Agricultural production system does not end with the actual harvesting of crops rather it extends to the postharvest system which primarily consists of crop storing, marketing, and transportation. However, temperature and humidity directly affect the quality of stored agricultural products. In a tropical country like the Philippines, tomato, lettuce, and other thin-skinned and highly moist crops degrade its quality and experience shape deformation over time. This study is a thematic taxonomy of intelligent postharvest storage systems discussing the techniques in the phenotyping of agricultural produce and emerging needs, trends in computer-vision-based postharvest systems, integration of artificial intelligence in postharvest systems, the current issues, challenges, and corresponding future directives in intelligent storage systems. Based on the systematic analysis, technical modeling of the storage system and postharvest crop quality grading are the emerging challenges in effectively storing crops for human consumption. It was found out that non-invasive high throughput methods for evaluation of quality and shelf life are needed. This can be done through vision-based fruit and vegetable quality grading and vision-based adaptive controls in the storage chamber. Overall, computer vision allied with artificial intelligence can make an intelligent postharvest storage system that is sustainable, profitable, and easy to implement.