Remote sensing image target detection method based on fusion convolution attention mechanism

The invention discloses a remote sensing image target detection method based on a fusion convolution attention mechanism, and solves the problems of low small target detection precision and slow convergence speed. The method comprises the steps of collecting and processing remote sensing image data; the method comprises the following steps of: constructing a feature extraction backbone network, and fusing a convolutional Transform encoding-decoding architecture; forming a fusion convolution attention mechanism target detection network model; and training and testing the target detection network model. According to the method, a downsampling feature extraction backbone network of a pyramid structure is adopted, and feature matrixes of the same size are output for input images of different sizes; a convolution module with deep convolution and point-by-point convolution is built, and the information extraction capability of local features of the remote sensing image is enhanced; a part of the attention head is r