Ming-Hui Cheng and Jyh-Horng Jeng
Video SuperResolution Using Classification ANN
1042 - 1045
2013
7
7
International Journal of Computer and Information Engineering
https://publications.waset.org/pdf/16533
https://publications.waset.org/vol/79
World Academy of Science, Engineering and Technology
In this study, a classificationbased video
superresolution method using artificial neural network (ANN) is
proposed to enhance lowresolution (LR) to highresolution (HR)
frames. The proposed method consists of four main steps
classification, motiontrace volume collection, temporal adjustment,
and ANN prediction. A classifier is designed based on the edge
properties of a pixel in the LR frame to identify the spatial information.
To exploit the spatiotemporal information, a motiontrace volume is
collected using motion estimation, which can eliminate unfathomable
object motion in the LR frames. In addition, temporal lateral process is
employed for volume adjustment to reduce unnecessary temporal
features. Finally, ANN is applied to each class to learn the complicated
spatiotemporal relationship between LR and HR frames. Simulation
results show that the proposed method successfully improves both
peak signaltonoise ratio and perceptual quality.
Open Science Index 79, 2013