The invention discloses an image rotation method based on FPGA, the method includes: taking the center of the image to be rotated as the origin (X ORIGIN , Y ORIGIN ) to establish an XY coordinate system, and the original image pixel coordinates (X 0 , Y 0 ) is divided into four quadrants, and the image rotation angle θ is divided into four situations, according to the original image pixel coordinates (X 0 , Y 0 ) is located in the quadrant and the different classifications of the image rotation angle θ to calculate the pixel coordinates (X, Y) after rotation, and write the data input by the video capture device at the corresponding address, and the FPGA is used to implement the above method. Compared with the traditional image rotation algorithm, many symbol operations and floating-point operations are more suitable for running in FPGA. While ensuring the real-time and smoothness of image rotation, it can be realized with only an entry-level FPGA. Compared with traditional GPU or DSP implementation methods, it reduces hardware costs and improves resource utilization efficiency.