| Summary: | 碩士 === 國立高雄應用科技大學 === 資訊工程系 === 106 === In recent years, deep learning has rapidly emerged, and all walks of life are transforming toward the artificial intelligence industry, and are widely used in image processing. Super Resolution is the use of low-resolution images to reconstruct corresponding high-resolution images. This technology is used in many places such as medical fields and monitor systems. The traditional method is to interpolate to fill in the information lost when the image is enlarged. The initial use of deep learning is SRCNN, which is divided into three steps, extracting image block features, feature nonlinear mapping and reconstruction. Both PSNR and SSIM have significant progress compared with traditional methods, but there are still some details in detail restoration. defect. SRGAN, proposed by C. Ledig, will generate anti-network applications to SR problems. The method is to improve the image magnification by more than 4 times, which is easy to produce too smooth. In this study, we hope to improve the EnhanceNet by training with different loss functions and different types of images to achieve better reconstruction results.
|
|---|
