Video optimization method and device, equipment and storage medium

The invention discloses a video optimization method and device, equipment and a storage medium, and relates to the technical field of video frames. The method comprises the following steps: adjusting a current video frame according to a motion feature and a previous global image feature of a previous video frame of the current video frame to obtain a first video frame; performing time dimension processing on the first video frame to obtain a second video frame; adjusting the current local image feature of the current local area of the second video frame to obtain a target video frame; and playing the target video frame. According to the technical scheme, in the video playing process, the video frame to be played currently is optimized, so that self-adaptive dynamic optimization of the video can be realized, and the user satisfaction can be improved. 本申请公开了一种视频优化方法、装置、设备及存储介质，涉及视频帧技术领域。该方法包括：根据当前视频帧的上一视频帧的运动特征和上一全局图像特征，对所述当前视频帧进行调整，得到第一视频帧；对所述第一视频帧进行时间维度处理，得到第二视频帧；对所述第二视频帧的当前局部区域的当前局部图像特征进行调整，得到目标视频帧；对所述目标视频帧进