Method for predicting position of artificial drifting fish collecting device or fish school

The invention discloses a method for predicting the position of an artificial drifting fish gathering device or a fish school. The method comprises the following steps: S1, acquiring a real-time screen capture image of a sea bird radar screen under the artificial drifting fish gathering device or the fish school; s2, extracting sea bird feature data in the real-time screenshot image through a computer vision technology, and establishing a sea bird radar image feature library under an artificial drifting fish gathering device or a fish school; s3, collecting sea bird radar image data in the sea bird radar in real time; s4, checking and correcting the sea bird radar image feature library by using the sea bird radar image data; s5, establishing a machine learning model; s6, utilizing the machine learning model to evaluate the possibility of the artificial drifting fish gathering device or fish school appearance; and S7, verifying the predicted artificial fish gathering device or fish school position, if the veri