IEEE Transactions on Computers
| Hi-index | 0.00 |
A two-level statistical classification procedure has been applied to the problem of detecting complex targets in aerial photography. At the first level, a set of classification functions designed on the basis of samples from the target class and from other images is used to make subdecisions on local-area statistically-designed features associated with the target class. At the second level these sub-decisions are combined into a single decision as to the presence or absence of the target. The nature of the data does not allow for the direct application of classical methods of multivariate discriminant analysis; rather, modifications of classical methods are used. This procedure has been simulated on a digital computer with the aid of a special input-output device which converts imagery to computer language. Excellent results were obtained on independent text samples of actual imagery.