Video demosaicing method based on auto-encoder and related device

The invention discloses a video demosaicing method based on an auto-encoder, and the method comprises the steps: carrying out the video frame extraction of a to-be-processed video, and obtaining a frame sequence; performing demosaicing processing on the frame sequence through an auto-encoder to obtain a processed frame sequence; wherein the auto-encoder is an auto-encoder obtained after training according to mosaic images and clean images; and connecting the processed frame sequences to obtain a target video. The frame sequence in the video is processed by adopting the auto-encoder, details inthe image are added, and the definition optimization effect is improved. The invention further discloses a video demosaicing system based on the auto-encoder, video demosaicing equipment based on theauto-encoder and a computer readable storage medium, which have the above beneficial effects. 本申请公开了一种基于自编码器的视频去马赛克方法，包括：对待处理视频进行视频帧提取处理，得到帧序列；通过自编码器对所述帧序列进行去马赛克处理，得到处理后的帧序列；其中，所述自编码器为根据马赛克图像和干净图像进行训练后得到的自编码器；将所述处理后的帧序列进行连接，得