Ship Detection by the RADARSAT SAR: Validation of Detection Model Predictions

Abstract

A statistical approach to point target detection in a clutter background is used to delineate the expected performance of the RADARSAT SAR (C-band HH polarization) for ship detection, and to compare the expected ship detection performance for the various RADARSAT SAR beam modes. Smaller ships may be detected with decreasing wind speed, increasing incidence angle, and finer resolution. ScanSAR Narrow Far is expected to be a good compromise between spatial coverage and detection probability. We present a quantitative validation of these model predictions based upon data acquired during a RADARSAT SAR ship detection/validation field program, which was held in March/April 1996 off the coast of Halifax, Nova Scotia. The program included buoy measurements of wind and wave conditions and RADARSAT SAR passes with known ships in each scene. We present validation of some of the model's key assumptions, in particular those concerning our hybrid C-band HH polarization ocean cross-section model, image probability density function, and ship radar signatures.

Use of Doppler Parameters for Ship Velocity Computation in SAR Images

“…Using (12) with the parameters listed in Table I, the expected slow time difference is 3.1 · 10 −2 s, which agrees with the results of the simulation presented in Fig. 3.…”

“…The ship is displaced in the positive azimuth direction, so it is moving toward the sensor. The resulting azimuth offset is 99 pixels, so according to (12), the radial velocity of the target can be estimated to be about −3.2 m/s, perfectly within the estimated range of values.…”

“…These ships were individuated by visual inspection looking at bright pixels over sea area, i.e., performing the same prescreening task as traditional adaptive threshold algorithms [9], [10]. For the 14 of these ships, the relevant wake had been also imaged, so the azimuth offset between the ship and the wake was used to derive an independent estimate of ship velocity by (12). The implemented algorithm to perform wakebased velocity estimation relies on the Radon transform [44] to extract the main wake features including the turbulent wake, which is used to compute ship-wake displacement.…”

“…The most common one is to try to guarantee that the probability of false alarm is constant over the image, thus leading to the definition of constant false alarm rate (CFAR) algorithms [9]. Relatively simple pdfs are used in CFAR algorithms, such as Gaussian, negative exponential, gamma, and K-distribution [12], which can be determined by a limited amount of parameters (e.g., mean, standard deviation, and other low-order moments). The practical implementation of these methods relies on nested boxes [10] that scan the image and are used to calculate locally the statistics of both the tested target and the surrounding background.…”

“…Plenty of experimental results are available about the use of adaptive threshold and CFAR algorithms for ship detection in SAR images [9], [10], [12]- [16]. The results clearly indicate that it is extremely difficult both to achieve a CFAR and to predict the false alarm rate theoretically.…”