Intelligent flower mushroom cultivation method based on behavioral cloning

The invention belongs to the field of practical application of a behavioral cloning technology, and provides an intelligent flower mushroom cultivation method based on behavioral cloning. The objective of the invention is to solve the ubiquitous problems that the cultivation process is difficult to separate from manual work and the efficiency is low when an intelligent plant factory is applied to the field of flower mushroom cultivation in the current flower mushroom cultivation technology. The main scheme comprises the following steps: segmenting a flower mushroom image by using a deep learning target detection algorithm; performing model training on the state information of the flower mushrooms and the corresponding expert behaviors by using a neural network; expanding the data set by using a data fusion mode; and applying the trained model to an intelligent system to realize expert behavior cloning. The method has the advantages that a YOLO-V5 target detection algorithm is adopted, the detection speed is h