Enhancement of the robustness on dynamic speckle laser numerical analysis

• The traditional normalization of the occurrence matrix presented some drawbacks.
• The proposed normalization presented lower dispersion when avoided the weighting.
• Tests with simulated and real data comproved the improvements.
• The change in the normalization increased the robustness of the analysis.

When a dynamic process occurs in a material under laser illumination the phenomenon that appears is named dynamic laser speckle, or biospeckle laser (BSL) if we have a biological material. The work with biological material and its dispersion of light brings considerable complexity, and the way we can deal with that complex outputs is based on a sophisticated analysis of the images associated to statistical approaches. One of the most known numerical analysis of the BSL has been applied in many applications, and it is named Inertia Moment, however its outputs have great coefficients of variation, most of the time attributed to the variability of the biological material. A change in the inertia moment method was done and the Absolute Value of the Differences (AVD) was presented as an alternative to reduce the variations and to follow a broader range of frequencies than before. However, it was not enough concerning with the variability of the outputs. This study aimed to improve the BSL technique in order to enhance the robustness of the numerical method known as Inertia Moment (IM) and improve the absolute value of the differences reducing even more its coefficient of variation by means of changes in the normalization provided in both methods. The new normalization was tested in simulated data, as well as in real data. The results showed the improvements of the methods, IM and AVD, with the reduction of the coefficients of variation of the activity in the outputs, increasing the robustness of the analysis.