Improvement in Spatial Resolution with HSI – SR and CNN

Abstract

Hyper spectral Super-Resolution (SR) is the method improves the spatial resolution of an image. A HIS SR is combined with the (CNN) model. It is used as a feature extractor. CNN contains various layers. These layers determine the color, edges and curves of the image. The process consists of three phases training, testing and implementing. Feature detection and its mapping are done by the convolutional layer. Pooling layer is another building block of CNN. Training is followed by testing and implementation. The algorithm is formed by using software called OpenCV using Python as a language. The hardware used for the testing phase is Raspberry pi. It is a compact size CPU.