Poultry house environment and poultry sensible temperature control system and method based on AI vision

The invention relates to the technical field of visual inspection, in particular to a poultry house environment and poultry sensible temperature control system and method based on AI vision. The method comprises the following steps: S1, collecting a complete period image of poultry growth from a farm, and taking the image as a sample; s2, making a data set, and dividing the collected image samples into a training set and a test set; s3, predicting the sensible temperature of the poultry according to mouth opening, wing spreading, compact spacing and dispersed distribution of the poultry so as to obtain estimation data; s4, the estimation data are uploaded to a microprocessor module, and the microprocessor module controls an environment control module to make corresponding adjustment; s5, the microprocessor module judges whether the environment temperature of the poultry house is within a specified temperature threshold value or not, and the temperature of the poultry house is regulated and controlled in time;