Weak rocks disintegration patterns recognition through image analysis

This paper presents results about image analysis application for identification and classification of disintegration patterns upon weak rocks, clay-bearing rocks and Intermediate Geomaterials IGMs. The research was conducted in order to obtain a reliable method for field and laboratory because upon these materials, it is quite difficult to acquire unaltered samples due to the quick disintegration after the materials are exposed to environmental conditions. Thus, the application of image analysis and color changes produced by disintegration advance has shown reliable results to be used as an alternative method to replace the traditional human eye based classification charts. Several images were taken upon different disintegration states and environmental conditions and were correlated with changes in color channels, using colorimetric indices and statistical image descriptors. As a result, a disintegration classification method based on Image entropy and color changes was stated, the findings were validated and compared with results from traditional methods using both natural and artificial samples with controlled disintegration levels, in addition, Hyperspectral images, were used as well.