A descriptor of speckle textures using box fractal dimension curve

• Published in: Optics and lasers in engineering 2018-07, Vol.106, p.47-55
• Main Authors: Rabal, H., Grumel, E., Cap, N., Buffarini, L., Trivi, M.
• Format: Article
• Language: English
• Subjects: Box fractal dimension; Dynamic speckle; Roughness; Speckle; Textures; Ultrasound images
• ISSN: 0143-8166; 1873-0302
• DOI: 10.1016/j.optlaseng.2018.02.006

•We propose a box fractal dimension algorithm that can be used in a wide set of applications, such as in incoherent illuminated ordinary images and in controlled, textures, static and dynamic speckle cases.•Generalization of the box fractal dimension in images by considering the curve obtained from its value as a function of the binarization threshold.•Results on the laser static speckle patterns in an example of roughness and on dynamic speckle quantitative measurements for free propagation geometry in controlled experimental conditions, in the evolution process of polymers (drying of paint) and in ultrasound speckle images.•Shown that the procedure to calculate the BFD can be extended to obtain a curve that is versatile and useful for the characterization of some aspects of the dynamics of speckle.•It was found that in some cases the BFDC could be approximated by a mathematical expression and an analogue to temperature used to characterize its time evolution. We propose a simple generalization of the box fractal dimension in images by considering the curve obtained from its value as a function of the binarization threshold. This curve can be used to partially describe ordinary images, textures, static and dynamic speckle patterns. We show some examples of different applications of this approach in some cases of interest.