Image enhancement method and device based on event signal neural coding mode

The invention discloses an image enhancement method and device based on an event signal neural coding mode, which combines the characteristics of a physical imaging model and a data driving mode, provides a robust and good coding mode, namely a neural event frame, for coding motion and time sequence information in an event signal flow, and improves the image enhancement efficiency through the neural event frame. A conversion relation between a low-quality image and a high-quality image is constructed, the problems of noise interference and the like in event signals are effectively solved, and deblurring and super-resolution of image signals are achieved. Meanwhile, a uniform high-speed video generation framework under the guidance of event signals is designed, and the recovery from low-speed videos to high-speed videos is realized. 本发明公开了一种基于事件信号神经编码方式的图像增强方法及装置，结合了物理成像模型和数据驱动两种方式的特性，提出了一种鲁棒的、良好编码事件信号流中运动和时序信息的编码方式--神经事件帧，通过神经事件帧，构建了低质量图像和高质量图像之间的转换关系，有效克服了事件信号中的噪声干扰等问题，实现了对于图像信号的去模糊以及超分辨率。同时，本发明设计了事件信号引导下统一的