Moreno Chávez, Gamaliel; Sarocchi, Damiano; Arce Santana, Edgar; Borselli, Lorenzo
Resumen:
The study of grain size distribution is fundamental for understanding sedimentological environments.
Through these analyses, clast erosion, transport and deposition processes can be interpreted and modeled. However, grain size distribution analysis can be difficult in some outcrops due to the number and
complexity of the arrangement of clasts and matrix and their physical size. Despite various technological
advances, it is almost impossible to get the full grain size distribution (blocks to sand grain size) with a
single method or instrument of analysis. For this reason development in this area continues to be fundamental. In recent years, various methods of particle size analysis by automatic image processing have
been developed, due to their potential advantages with respect to classical ones; speed and final detailed
content of information (virtually for each analyzed particle). In this framework, we have developed a
novel algorithm and software for grain size distribution analysis, based on color image segmentation
using an entropy-controlled quadratic Markov measure field algorithm and the Rosiwal method for
counting intersections between clast and linear transects in the images. We test the novel algorithm in
different sedimentary deposit types from 14 varieties of sedimentological environments. The results of
the new algorithm were compared with grain counts performed manually by the same Rosiwal methods
applied by experts. The new algorithm has the same accuracy as a classical manual count process, but the
application of this innovative methodology is much easier and dramatically less time-consuming. The
final productivity of the new software for analysis of clasts deposits after recording field outcrop images
can be increased significantly.
Descripción:
The study of grain size distribution is fundamental for understanding sedimentological environments.
Through these analyses, clast erosion, transport and deposition processes can be interpreted and modeled. However, grain size distribution analysis can be difficult in some outcrops due to the number and
complexity of the arrangement of clasts and matrix and their physical size. Despite various technological
advances, it is almost impossible to get the full grain size distribution (blocks to sand grain size) with a
single method or instrument of analysis. For this reason development in this area continues to be fundamental. In recent years, various methods of particle size analysis by automatic image processing have
been developed, due to their potential advantages with respect to classical ones; speed and final detailed
content of information (virtually for each analyzed particle). In this framework, we have developed a
novel algorithm and software for grain size distribution analysis, based on color image segmentation
using an entropy-controlled quadratic Markov measure field algorithm and the Rosiwal method for
counting intersections between clast and linear transects in the images. We test the novel algorithm in
different sedimentary deposit types from 14 varieties of sedimentological environments. The results of
the new algorithm were compared with grain counts performed manually by the same Rosiwal methods
applied by experts. The new algorithm has the same accuracy as a classical manual count process, but the
application of this innovative methodology is much easier and dramatically less time-consuming. The
final productivity of the new software for analysis of clasts deposits after recording field outcrop images
can be increased significantly.