Genetic algorithms have been widely used as an
optimization technique for search space. In this study, the
genetic algorithm is integrated with the DCL algorithm to
improve the unsupervised training results for the
classification of remotely sensed data. The simulation
results and their comparison with the K-means algorithm,
the genetic K-means algorithm and the DCL algorithm will
be demonstrated for unsupervised training. Like the
conventional K-means clustering, this new approach can
be applied in many different application areas such as
computer vision, pattern classification, industrial products
inspection, . . . , etc. This computation-oriented algorithm
is suitable for parallel implementation. The evaluation of
the J-M distance for the effectiveness measure and
classification accuracy assessment are under investigation.