Lacrimal discharge detection model construction method, lacrimal discharge detection method, lacrimal discharge detection device and lacrimal discharge detection equipment

The invention discloses a tear overflow detection model construction method, a tear overflow detection method, a tear overflow detection device and tear overflow detection equipment, which are applied to the field of artificial intelligence, and the method comprises the following steps: obtaining an ocular surface image of a tear river dyed by fluorescein sodium as sample data; training an image quality screening model by using the qualified and unqualified sample data to obtain a trained image quality screening model; training the tear overflow degree preliminary diagnosis model by using the normal, light and medium tear overflow and heavy tear overflow sample data to obtain a trained tear overflow degree preliminary diagnosis model; training a light-moderate dichotomy model by using the sample data of moderate tear overflow and mild tear overflow to obtain a trained light-moderate dichotomy model; the trained image quality screening model, the trained tear overflow degree preliminary diagnosis model and the