Single Frame Super Resolution using Reconstruction Method

Fumiya Sadoyama, Chikatoshi Yamada, Shuichi Ichikawa


High definition technique is an image processing which is often constructed by using bicubic interpolation. How- ever, such as the bicubic interpolation that is a traditional method has problems. For examples, ringing and jaggies are caused without the high frequency components in image processing. In this paper, we propose a single frame super resolution using reconstruction method. Therefore, in order to reduce their problems, the proposed method consists of four parts: (1) detection of similar structures, (2) data fusion, (3) bilateral filter using local variance, and (4) bilateral back projection using local variance. Finally, we demonstrate experimental results of super resolution for some images. Moreover, we show that the proposed method is more effective in comparison with traditional method.


Single Frame; Super Resolution; Similar Structure; Bilateral Back Projection; Local Variance

Full Text:



  • There are currently no refbacks.