One of the major problems encountered in developing and testing any target classification scheme is the collection of an adequate data base for training and testing the classifier. This is particularly a problem if one must obtain radar information from a large number of viewing angles and over a large number of targets, some of which may be unfriendly. A method for obtaining this information by simulation is described. Synthetic aperture radar signals such as would be obtained from either a moving or a stationary ship, are simulated and it is shown how such a facility can be used in the development of an automatic ship classification system. The images of stationary ships are used to train and test the classification system and the images of moving vessels are used to show how the classifier performance degrades if the motion is left uncorrected. A readily available data base such as Jane's Book of Ships is used to compile 3-D outlines of the vessels of interest. These outlines, accurate down to the 1-m level, are used to decompose the ship's superstructure into a set of radar scatterers. The returns from the scatterers are then summed, taking into account the viewing angle of the radar. Using this technique, the radar scattering models for a complete library of ships can be formed, and the library used to develop and test various automatic classification schemes. Examples of simulated SAR images of ships are given and the real image of a particular ship is compared to its simulated one. The limitations of the technique are discussed, and the use of the data in testing an actual classification scheme is briefly outlined.


    Access

    Access via TIB

    Check availability in my library


    Export, share and cite



    Title :

    Use of Simulated Synthetic-Aperture Radar: Signals of Ships in the Development of an Automatic Ship Classification System


    Contributors:

    Publication date :

    1986


    Size :

    1 pages


    Type of media :

    Report


    Type of material :

    No indication


    Language :

    English