Egg classification system and method based on machine vision

The invention discloses an egg classification system and method based on machine vision. The system comprises an image collecting device, a light source, an infrared sensing device, a control device,a classification device and a computer. The method includes: performing image preprocessing including double-industrial-camera image collecting, smoothing, grey level transformation and the like; using an improved maximum between-class variance method to improve Canny operator edge detection, extracting the features such as height, shape and area of yolk and egg white which is beaten by an eggbeater, calculating a Haugh value, and building a freshness feature knowledge base based on the neural network; performing egg freshness matching and classification according to the freshness feature knowledge base. The egg classification system and method has the advantages that the machine vision technology is utilized, and the system is high in running speed, capable of achieving real-time egg freshness classification, hig