In High-Frequency Surface Wave Radar applications (HFSWR), targets at close range are often masked by very strong sea clutter returns in the range-Doppler spectrum. The clutter is highly non-homogeneous in the Doppler dimension, presenting two peaks at resonance Doppler frequencies, referred to as Bragg lines, and several smaller side peaks. Due to this complex scattering mechanism, a Gaussian assumption for clutter returns does not hold, and several works propose modeling sea clutter returns in HF range-Doppler spectra as a Weibull distribution. This work presents an analysis of the performance of a cell-averaging constant false alarm rate (CA-CFAR) algorithm designed for Weibull-distributed clutter. A closed-form probability of detection of the Weibull CACFAR is compared to a numerical simulation of sea clutter based on a physical model of sea radar cross-section (RCS). The clutter model takes into consideration wind conditions, as well as the operating parameters of the radar. It is demonstrated through numerical simulation of the physical model that the clutter echo distribution, depending on the sampling position in the range-Doppler spectrum and proximity to a Bragg line, can take the form of an exponential or a Rayleigh distribution. Thus, the overall distribution of clutter returns can be represented by a Weibull model. Results indicate that the closed-form analytical expression act as an upper bound for detector performance, that is, in practice, degraded by the strong peaks of the clutter power.
This article presents a deep neural network‐based constant false alarm rate (NNB‐CFAR) detector for simulated high‐frequency surface wave radar (HFSWR) data. A deep neural network is trained to identify fluctuation parameters of each cell of a range‐Doppler power spectrum based on the patterns present in the neighbouring cells. The estimated parameters are then used for calculating a detection threshold with a user‐specified probability of false alarm. To train the network, a realistic model of HFSWR echoes is used for generating a large labelled range‐Doppler image dataset, including many possible clutter scenarios and interfering target echoes. Several CFAR windows are extracted from the training range‐Doppler dataset and used as training data. The neural network is trained to replicate the output of a maximum likelihood estimator based on the reference cells of the CFAR window. The NNB‐CFAR algorithm was then compared to traditional CFAR algorithms by identifying targets in the second set of simulated range‐Doppler images. The probability of detection was also experimentally measured in the context of HFSWR for all algorithms. Results show that the technique can significantly improve detection rates amid strong clutter.
This paper presents a study regarding the impact of using imperfect reference images containing targets and artifacts in the performance of change detection methods. The presented analysis uses a change detection method based on Bayes' Theorem recently proposed. The experimental evaluation is carried out using wavelength-resolution SAR images obtained using the CARABAS II SAR system. The experimental setup considers two types of reference images, i.e., imperfect and ground scene prediction (GSP) generated images. The imperfect images are those available in the dataset. The GSP-generated images are obtained by the GSP method and tend not to contain targets and artifacts. Results indicate that the use of reference images obtained by the GSP method provides a false alarm reduction in the evaluated scenarios when compared with the CDA implementation with imperfect reference images. For instance, a FAR reduction from 0.667 to 0.229 is observed in an evaluated setup.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.