Application of Spectral Methods for Synthesis of Symptoms for Qualification of Agricultural Products

Abstract

The basic stages during automatic qualification of agricultural products are: acquisition of initial information of the quality state, mathematical proceeding of this information and decision making for object qualification to defined sets according to the state of quality. One of the steps in the second stage is transformation the patterns of qualified products to a new space of symptoms. In this paper an approach for transformation the initial description of the objects to a new space of spectral symptoms, independently of the size of the initial description of the object is presented. A relation for synthesis of symptoms for recognition has been generated.