Publication

Abstract : Hyperspectral imaging is a technique in which the information is collecting and processing across an electromagnetic spectrum. Fractal Image Processing deals with various dimensions of an image. Hyperspectral analysis are applied on multiple frequency bands of images, and produces spatial and spectral information. HSI can elongate 390 to 700 nm i.e UV to infrared and near-infrared wavelength areas. This technique based system engenders narrow band(<; 390 to 700 nm>) images of different wavelengths. By utilizing HIS, visualization can be elongated to invisible wavelengths. Hyperspectral based techniques have applications in various fields such as medical diagnosis, agriculture, food processing, remote sensing etc. In this project, classification of hyperspectral images is considered based on Neural Network. Present paper also discusses the conventional hyperspectral image classification methods, their comparisons and impact on accuracy of classification.