Feature generation from digital images using pseudo-fractal algorithm and its four modifications

The main plot of the paper is to present the authors’ original method of feature generation from digital images and to report on comparison of five various algorithms, which realise the method. The algorithms are based on the same authors’ idea, consists in quantitative description of similarity intensity between various parts of an image in various scales. To do so the algorithms take advantage of fractal coding based on an Iterated Function System. Therefore generated features can rightly be called similarity features. In this paper we show that similarity features, when combined with other well known ones, can improve recognition results in some image classification tasks. After the presentation of the algorithm way of working, we compare their properties and report classification results obtained in two different pattern recognition experiments. Moreover, the paper contains a discussion of obtained results, and possible future applications of similarity features.