Image splicing method and device, model training method and device, electronic equipment and medium

The invention provides an image splicing method and device, a model training method and device, electronic equipment and a medium. The method comprises the steps of obtaining a to-be-spliced image sequence; the image sequence is input into a trained neural network model to output a perspective transformation matrix corresponding to the image sequence and the offset of each image frame, the perspective transformation matrix is used for describing the transformation relation of the image in perspective transformation, and the offset is used for adjusting the position of each image frame in the splicing process; and according to the perspective transformation matrix and the offset, splicing each image frame in the image sequence to generate a final spliced image. According to the method and the device, the problem of image splicing of the scanning pen is effectively solved, and the accuracy, consistency and stability of a splicing result are improved. 本申请提供一种图像拼接方法、模型训练方法、装置、电子设备及介质，所述方法包括：获取待拼接的图像序列；将所述图像序列输入训练