Superpixel Boundary based Image Upsampling
| Autor: | Yun-Yang Lai, 賴昀揚 |
|---|---|
| Rok vydání: | 2018 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 106 In this paper, we propose a boundary preserved image upsampling method based on superpixel boundary information. We design a novel image upsampling interpolation function, which can combine superpixel boundaries to enhance the edges of the upsampled results. By using the low-resolution image and the corresponding high-resolution image’s superpixel boundary information, our method can produce the high-resolution image. In order to obtain the high-resolution image’s superpixel boundary information, we train a convolution neural network model to produce the proper superpixel segmentation from the low-resolution image. With the interpolation function and the superpixel boundary information obtained directly for the low-resolution image, our method suits the image upsampling problem. Our method can produce sharp edges while the traditional methods can only produce blurry edges. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|