21 Target Detection in Hyperspectral Imaging Using Neural Networks
Edisanter Lo and Emmett Ientilucci
21.1 Introduction
Artificial neural networks have worked well in hyperspectral imaging for classification but have difficulty in hyperspectral imaging for target detection of rare targets. The difficulty is caused by the scarcity of available target pixels in the image for training and validating the neural networks. Conventional stochastic gradient descent algorithms for neural networks use a fixed stepsize that has to be predetermined by trial and error to achieve convergence. The first objective of this chapter is to develop a gradient descent algorithm for solving the stochastic optimization problem of a ...
Get Signal and Image Processing for Remote Sensing, 3rd Edition now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
