Classification of User Interface Elements by Calculating Variance

The image classification has been a difficult problem to deal in the last decades, since it requires in the most of cases the application of complex and sophisticated algorithms as well as large computational resources. In this paper we propose a series of techniques for the classification of different elements used in the creation of user interfaces, which do not require large computational resources so they are quick and can be parallelized. These techniques consist of a) isolating the image by extracting the area that contains the sketches of user interface elements made by freehand eliminating areas that do not contain information, b) obtaining representative areas from histograms that represent the sketches, c) standardize the information contained in the histograms, d) calculate the variance of the information of the histograms and e) disassociate this information so that finally be used to carry out a classification. The application of this set of techniques gives us a quick result to classify sketches of user interfaces.