改进的自适应冲击滤波图像超分辨率插值算法

Interpolation Algorithm Based on Improved Adaptive Shock Filter in Image Super-Rresolution

针对基于插值的图像超分辨算法造成图像边缘扩散和引入噪声的缺点,该文提出一种改进的自适应冲击滤波模型的图像超分辨率插值方法.根据插值后的初始高分辨率图像梯度自适应调整冲击滤波的权重,针对不同的图像边缘相应地减少图像的边缘扩散,同时利用前向扩散消除噪声.相对传统冲击滤波方法,避免了锯齿、块效应等人工痕迹,有效的保留了细节纹理特征.理论分析和实验结果表明,文中方法获得更好的超分辨率结果,主观效果得到明显改善,客观指标得到一定提高.

In view of the disadvantages such as edge diffusion and introducing noise in image super-resolution based on interpolation algorithm,an interpolation method based on improved adaptive shock filter is proposed in this paper.The weighting of shock filter can be adjusted according to the gradient of initial high-resolution image.Edge diffusion is reduced for different image edges.The noise is removed by forward diffusion.The proposed model preserves texture features efficiently and overcomes the artifact such as jaggies and blocking effects compared with traditional shock filter.Theoretical analysis and experiments demonstrate that the method proposed in the paper has a better visual result.The subjective quality is improved obviously,and the objective indicators are improved partly.