Acta Montanistica Slovaca (Jun 2004)
Object Recognition in System FOTOM
Abstract
The recognition of points of interest leads to the detection and consequent specification of the values of Xí and Yi coordinates. The result is a data file of the coordinates. The problem domain is divided into two parts.The first part of the paper deals with the basic principles of figure segmentation. The main goal of the segmentation is to split the figure into parts. whjch closely cohere with the object or domains of real world. The result should be pictures of real objects located on the figure. Thresholding methods, mathematic morphology. Brightness attributes and others are mentioned here. The detection of points of interests on the light track and the determination of points of interest using neural networks are discussed at the end of the part.The second part concentrates ori the completion of the module FOTOM 1 by new functions. The design and realizaton of the module FOTOM7 is described too. These modules implemen the detection of points of interests on the light track and the determination of points of interest using neural networks. The experience gained during the implementation of the module FOTOM 7 are presented at the end.
