Power Law Dependencies to Detect Regions of Interest

This paper presents a novel approach to detect regions of interest in digital photographic grayscale images using power laws. The method is intended to find regions of interest in various types of unknown images. Either Zipf law or inverse Zipf law are used to achieve this detection. The detection method consists in dividing the image in several sub-images, computing the frequency of occurence of each different image pattern, representing this distribution by a power law model and classifying the sub-frames according to the power law characteristics. Both power laws models allow region of interest detection, however inverse Zipf law has better performances than Zipf law. The detection results are generally consistent with the human perception of regions of interest.

[1]  E. B. Newman,et al.  Tests of a statistical explanation of the rank-frequency relation for words in written English. , 1958, American Journal of Psychology.

[2]  Michael Brady,et al.  Saliency, Scale and Image Description , 2001, International Journal of Computer Vision.

[3]  N. Vincent,et al.  Compressed image quality and Zipf law , 2000, WCC 2000 - ICSP 2000. 2000 5th International Conference on Signal Processing Proceedings. 16th World Computer Congress 2000.

[4]  George Kingsley Zipf,et al.  Human behavior and the principle of least effort , 1949 .

[5]  Huberman,et al.  Strong regularities in world wide web surfing , 1998, Science.

[6]  Nicole Vincent,et al.  A method for detecting artificial objects in natural environments , 2002, Object recognition supported by user interaction for service robots.

[7]  G. Zipf The Psycho-Biology Of Language: AN INTRODUCTION TO DYNAMIC PHILOLOGY , 1999 .

[8]  Rosario N. Mantegna,et al.  Numerical Analysis of Word Frequencies in Artificial and Natural Language Texts , 1997 .

[9]  H E Stanley,et al.  Linguistic features of noncoding DNA sequences. , 1994, Physical review letters.

[10]  Li Fan,et al.  Web caching and Zipf-like distributions: evidence and implications , 1999, IEEE INFOCOM '99. Conference on Computer Communications. Proceedings. Eighteenth Annual Joint Conference of the IEEE Computer and Communications Societies. The Future is Now (Cat. No.99CH36320).

[11]  James Ze Wang,et al.  Unsupervised Multiresolution Segmentation for Images with Low Depth of Field , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[12]  H. Stanley,et al.  Modelling urban growth patterns , 1995, Nature.

[13]  B. M. Hill,et al.  Zipf's Law and Prior Distributions for the Composition of a Population , 1970 .