On the importance of using knowledge and the manner of incorporating it into the image analysis

During the last 3 decades, many remote sensing image analysis methods have been proposed by the researchers including the statistical approaches, knowledge based and model based algorithms. Beside the accuracy improvement that has been reached by these methods, it has been approved that although this is important to make use of the knowledge about the objects and processes but the most important thing is how to use these knowledge in the image analysis. In this paper we are going to show that using knowledge of land cover objects and processes in a proper manner can improve efficiently the accuracy of the image analysis. During this study, some aspects of the traditional maximum likelihood classifier are examined and a method for incorporating knowledge of objects and processes into the image analysis is introduced

کلیدواژه ها:

Knowledge ، Model Based Image Analysis ، Maximum likelihood Classification ، Remote Sensing