Decomposition of biospeckle signals through granulometric size distribution.

We present a method for the analysis of dynamic speckle signals based on morphological granulometry. We obtain selected information differentiating the morphological patterns of the temporal history of each pixel through the granulometric size distribution. The method is exemplified by detecting bruised regions on apples and studying the germination of corn seeds. Different levels of activity are observed in the bruised and healthy areas of the apple, within a certain range of the morphological spectrum. Likewise, the activity of the corn seed embryo can also be distinguished from the endosperm area.