Super Resolution: Reconstruction of High-resolution Image from a Set of Low-resolution Images

Although imaging sensors are the dominant technologies for both user and industry applications, they still have several physical limitations, such as noise and limited spatial resolution. These limitations can be overcome, based on device electronics and physics technology. However, a promising solution is a signal processing approach that has been one of the most active research areas, and it is called Super Resolution (SR). This work proposes SR algorithm that uses an affine block-based with the Maximum Likelihood. A number of experiments were performed with the proposed system to obtain reconstructed High Resolution (HR) images of different resolutions from the same set of Low Resolution (LR) images. Also, a number of experiments were performed to evaluate its behavior as a function of the number of available LR images. The algorithm improves the accuracy of translational registration and accurately recovers HR image even in the case where just very a few input images are provided. This work should be especially useful to professionals in Image Processing, Signal Processing, and Electronics fields, or anyone else who may be considering utilizing Resolution Enhancement.