ISAR-Image recognition using optimized HONN by a Metaheuristic algorithm

Abstract

In the field of ISAR Data extraction, many drawbacks appear and make this field a very challenging one. Therefore, using higher order neural networks (HONN) became an useful way to cope with problems like the inability to scale with the complexity of the problem and the sluggish converge rate that results in a lengthy training period. The Functional Link Artificial Neural Network (FLANN), a higher order neural network, was optimized in this paper using a revolutionary metaheuristic inspired by the Firefly algorithm to identify radar targets. Results from experiments demonstrate the effectiveness of the training method.