Dynamic speckle texture processing using averaged dimensions

Dynamic speckle or biospeckle is a phenomenon generated by laser light scattering in biological tissues. It is also present in some industrial processes where the surfaces exhibit some kind of activity. There are several methods to characterize the dynamic speckle pattern activity. For quantitative measurements, the Inertia Moment of the co occurrence matrix of the temporal history of the speckle pattern (THSP) is usually used. In this work we propose the use of average dimensions (AD) for quantitative classifications of textures of THSP images corresponding to different stages of the sample. The AD method was tested in an experiment with the drying of paint, a non biological phenomenon that we usually use 14 as dynamic speckle initial test. We have chosen this phenomenon because its activity can be followed in a relatively simple way by gravimetric measures and because its behaviour is rather predictable. Also, the AD was applied to numerically simulated THSP images and the performance was compared with other quantitative method. Experiments with biological samples are currently under development.