Pictures classification in multimedia databases
Tomasz Gąciarz, Krzysztof Czajkowski · 2010
This paper concerns pictures fast classification on the basis of information included in them. Features as occurrence of straight lines, average brightness in particular areas of picture, text detection as well as human face occurrence are taken into consideration. On the basis of these features, images are classified to one of several categories, by using neural network. Chapter describes implementation and results obtained using mentioned approach.