Non-Cooperative Target Recognition by Means of Singular Value Decomposition Applied to Radar High Resolution Range Profiles

Abstract

Radar high resolution range profiles are widely used among the target recognition community for the detection and identification of flying targets. In this paper, singular value decomposition is applied to extract the relevant information and to model each aircraft as a subspace. The identification algorithm is based on angle between subspaces and takes place in a transformed domain. In order to have a wide database of radar signatures and evaluate the performance, simulated range profiles are used as the recognition database while the test samples comprise data of actual range profiles collected in a measurement campaign. Thanks to the modeling of aircraft as subspaces only the valuable information of each target is used in the recognition process. Thus, one of the main advantages of using singular value decomposition, is that it helps to overcome the notable dissimilarities found in the shape and signal-to-noise ratio between actual and simulated profiles due to their difference in nature. Despite these differences, the recognition rates obtained with the algorithm are quite promising.

Non‐cooperative identification of civil aircraft using a generalised mutual subspace method

“…Table 3 shows the comparative recognition results obtained for the aforementioned methods using the same test and training sets presented in this paper. The parameters selected for each algorithm are equal to the ones chosen in their respective references, that is for entry F2 in the Table, the threshold parameter for the subspace design is set to (77 = 0.85) [30] and for entry F3 the threshold for the feature subspace formation is set to (77 = 0.99) as [4] states. In the case of algorithm MSM the threshold is set to (77 =0.85) due to the similarity between this algorithm and F2.…”

“…In order to further assess the performance, the method is compared to three different algorithms: the first one is the MSM already described in Section 2.1; the second one is a subspace-based method presented in [30] where SVD is used to define the subspaces that will represent each target (with a threshold of 85%) and the algorithm used for identification is also based on angle between subspaces, although in this specific case, between a vector and a subspace with a weighting element; lastly, another subspace-based method presented in [4] is used for comparison, the PCA-based minimum reconstruction error approximation, where PCA is used to extract the feature subspace of a frame of HRRPs (with a threshold of 99%) and then, the algorithm decides the test sample class considering its minimum reconstruction error in the feature subspaces. The reader is referred to the cited papers for more information about the algorithms.…”

Automatic Acoustic Target Detecting and Tracking on the Azimuth Recording Diagram with Image Processing Methods

“…where φ D is the azimuth satisfying the constraint in Equation (25) and θ is the ending azimuth of the simulated trajectory of the optimal template at (φ D , t).…”

“…The underwater acoustic target azimuth recording diagram is essentially a fusion of useful information and interference noise, so it can be projected into useful information and interference noise subspaces. This paper enhances the azimuth recording diagram using principal component analysis (PCA), which has been used in the sonar or radar systems [21][22][23][24][25].…”

Singular Value Decomposition Applied to Non-Cooperative Target Identification

“…Radar recognition has received intensive attention from the NCTI community, since it permits to recognize targets at long distance and under poor visibility conditions without the need of communication with targets and even with them being unaware of it [2]. Non Cooperative Target Identification (NCTI) by radar relies on a comparison between the measured target signatures and a reference database [3][4][5]. This can carried out by different techniques, including Jet Engine Modulation (JEM), HRRPs and Inverse Synthetic Aperture Radar (ISAR) images.…”