Method for constructing colony count regression prediction model and method for detecting freshness of fish by using model

The invention relates to a method for constructing a colony count regression prediction model and a method for detecting the freshness of fish by using the model. The construction method of the colonycount regression prediction model comprises the steps of performing hyperspectral image detection on a fish sample, preprocessing a hyperspectral image, obtaining a training sample and establishing the colony count regression prediction model; after the colony count regression prediction model is constructed, acquiring a hyperspectral image of the to-be-detected fish sample through a hyperspectral sorter, and predicting the colony count of the to-be-detected fish sample so as to determine the freshness of the to-be-detected fish sample. According to the invention, a kernel extreme learning machine is optimized by using the differential evolution algorithm to further establish the fish colony count regression model; according to the model, nondestructive detection can be carried out on thecolony count content of t