Automated Classification of Bitmap Images using Decision Trees
Pavel Surynek, Ivana Lukšová · Polibits · 2011
"The paper addresses the design of a method forautomated classification of bitmap images into classes describedby the user in natural language. Examples of such naturallydefined classes are images depicting buildings, landscape, artisticimages, etc. The proposed classification method is based on theextraction of suitable attributes from a bitmap image such ascontrast, histogram, the occurrence of straight lines, etc.Extracted attributes are subsequently processed by a decisiontree which has been trained in advance. A performedexperimental evaluation with 5 classification classes showed thatthe proposed method has the accuracy of 75%-85%. The designof the method is general enough to allow the extension of the setof classification classes as well as the number of extractedattributes to increase the accuracy of classification."