Research on radar target recognition by fusion fuzzy logic algorithm

Abstract

For reliable identification and categorization of observed objects in radar systems, radar target recognition is a crucial job. In this study, we provide a method for identifying radar targets that uses a fusion fuzzy logic algorithm to increase identification precision. To deal with the uncertainties and imprecisions involved in target detection, the system combines input from several radar sensors and uses fuzzy logic methods. Inverse Synthetic Aperture Radar (ISAR) image-based classifiers are often combined or fused to examine complementary information. As a consequence, the findings from each classifier will be merged to increase the overall recognition rate. For this purpose, fusion methods are primarily used by automatic target recognition systems. This strategy is one of the most important current areas in target recognition research due to the empirical proof of its efficacy. The recognition combination will be described in this research utilizing fuzzy fusion based on three classifiers: Logistic Regression (LR), Artificial Neural Network (ANN), and Deep Convolutional Neural Network (DCNN) classifiers. To accomplish this goal, we have been using the Mamdani and Sugeno models. We have used an ISAR image database that was rebuilt from an anechoic chamber to increase the effectiveness of the suggested technique. It is intended to display every single result obtained from every individual classifier as well as the aggregated results. The proposed model has provided an accuracy and recognition rate of 97% and 94.6% respectively.