基于受限玻尔兹曼机的受损图像还原方法研究

Research on Damaged Image Restoration Based on Restricted Boltzmann Machine

根据受限玻尔兹曼机(RBM)的特点,提出基于受限玻尔兹曼机的图像修补方法,并在MNIST数据集上对模型进行验证.结果表明,采用该图像修补算法能够提高图像修复质量.

According to the characteristics of the RBM,this paper proposes an image inpainting method based on the RBM,and verifies the model on the MNIST dataset.The results show that the image inpainting algorithm used in this paper can improve the quality of image inpainting.