Shape-based recognition of targets in synthetic aperture radar images using elliptical Fourier descriptors

Abstract

This paper primarily investigates the use of shape-based features by an Automatic Target Recognition (ATR) system to classify various types of targets in Synthetic Aperture Radar (SAR) images. In specific, shapes of target outlines are represented via Elliptical Fourier Descriptors (EFDs), which, in turn, are utilized as recognition features. According to the proposed ATR approach, a segmentation stage first isolates the target region from shadow and ground clutter via a sequence of fast thresholding and morphological operations. Next, a number of EFDs are computed that can sufficiently describe the salient characteristics of the target outline. Finally, a classification stage based on an ensemble of Support Vector Machines identifies the target with the appropriate class label. In order to experimentally illustrate the merit of the proposed approach, SAR intensity images from the well-known Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset were used as 10-class and 3-class recognition problems. Furthermore, comparisons were drawn in terms of classification performance and computational complexity to other successful methods discussed in the literature, such as template matching methods. The obtained results portray that only a very limited amount of EFDs are required to achieve recognition rates that are competitive to well-established approaches.

Обработка Сигналов От Пространственно Распределённых Целей

“…Метод опорных векторов позволяет достигнуть точности 90 % и выше в задачах распознавания объектов базы MSTAR [36,37,190,194] в сочетании со сравнительно низким уровнем ошибок пропуска цели [36,39,134,140,190,195]. Метод опорных векторов успешно используется в сочетании с различными приемами преобразованию пространства признаков, такими как метод главных компонент [36,37] и метод независимых компонент [44], неотрицательного матричного разложения векторов признаков обучающей выборки (NMF) [134], а также выделению дополнительных дешифровочных признаков с помощью Фурье-преобразования [194] и нахождения центральных моментов изображения [119,190].…”

“…Некоторого ускорения расчетной процедуры можно достичь, используя низкоранговое матричное разложение [48,49,51]. Точность классификации РЛИ возрастает обратно пропорционально количеству распознаваемых классов [39,194], поэтому использование метода SVM не является универсальным решением. Отдельно следует отметить необходимость предварительного анализа исходных данных, необходимых для грамотного выбора адекватных параметров ядра классификатора [39,193], что полностью предопределяет результаты работы алгоритма.…”

Analyzing the Dynamics of Lung Cancer Imaging Data Using Refined Fuzzy Entropy Methods by Extracting Different Features

“…Kuhl and Giardina [64] in 1982 introduced EFDs features to classify the solid objects such as box and car etc. Moreover, in pattern recognition systems [65], [66], EFDs features are most widely used. EFD method made up of an expansion of a shape as sum of ellipse.…”

Automatic target recognition scheme for a high-resolution and large-scale synthetic aperture radar image

“…In fact, since the Massachusetts Institute of Technology Lincoln Laboratory developed the SAR ATR system 39 for the defense advanced research projects agency-sponsored semi-automated imagery intelligence processing program in 1997, many methods 1,9,14,17,19,21,37,38,40 for an SAR ATR framework have come into being. However, most of those subsystems above were designed without considering the total integration of the SAR ATR; they were only for one part (detection, discrimination, or classification) of an SAR ATR system.…”