In this study, a novel multimodal machine vision system utilizing color, 3-D depth, spectral, and X-ray imaging technologies was designed for quality inspection of onions. A LabVIEW program was also developed to control and synchronize hardware devices for data acquisition. The system can nondestructively acquire comprehensive information about onions in color, 3-D, spectral, and X-ray domains. The proposed multimodal system demonstrated an enhanced sensing capability compared to conventional classification systems for fruits and vegetables. By integrating the cameras and sensors with complementary sensing capabilities into one system, key quality parameters of onions were accurately assessed with one scan. In the future, the proposed multisensor-based machine vision system can be used to measure additional onion quality factors, and to grade onions using more comprehensive criteria. The presented system and methods can also be potentially extended for postharvest quality inspection of other agricultural products.
