Multi-Spectral Camera System to Improve Food and Material Identification

Visible cameras have been widely used in many places for food and material identification, such as food sorting and produce labeling. Visible camera sensors, working in the 400nm-700nm range, can identify food with different colors, shapes, or other visible characteristics that are otherwise unidentifiable to visible cameras. Visible image sensors struggle to identify food and material with similar color, as they are limited to the colors ranges human eyes can sense[1] . In this study, a Sony IMX991 sensor was used to capture both visible and infrared images. It was demonstrated that some materials can be distinguished with either visible or infrared narrow band images[2] . Material identification accuracy can be improved by analyzing both visible and infrared images. A machine learning algorithm can be developed by analyzing all images captured under different conditions, accuracy of food and material identification can be significantly improved with continuous image process algorithm development and machine learning training.