Robust aircraft classification using moment invariants, neural network, and split inversion learning

Summary form only given, as follows. A robust approach for classifying aircraft in the presence of noise was proposed and simulated. Preprocessing provides constant moment invariants for images in which the object is translated in position, rotated, or changed in scale. In the presence of noise, the moments are shown to be rotation-invariant. These moment invariants, computed for different levels of noise, are used to train a neural network to identify the aircraft. It is shown that training is very much faster for a split inversion algorithm than for back propagation. The resulting network provides accurate classification in high levels of noise