Dr. Vijayan Asari, DirectorPhone: 937-229-4504
Single Image Super Resolution
Super-Resolution enhancement refers to the process of obtaining high resolution image from a single low resolution image. This nonlinear interpolation process produces sharper images with more unique features. The flowchart below describes the super resolution technique.
In the images below, we illustrated the super resolution capabilities of creating a clearer images. By doubling the height and width of the image, for instance NxM images become 2Nx2M image, we increase the image resolution by four times that of the original.
Super resolution can be used in Scene Analysis and Understanding researches, such as 3D Scene Creation, by providing algorithms more key feature points to be extracted from the original scene. In the figures below, observe the increase in sharpness as super resolution is applied.
Super resolution also plays an essential role in Wide Area Surveillance Object and Pedestrian Tracking and Object Classification. The images below illustrate the enhancement super resolution provides in low resolution images of small objects.
<< Back to Research Projects