The Geo-Statistical Modeling of Earth Color Properties Using Core Images Processing in RGB System for Separation of Copper Alteration (Case Study: Kohang and Meiduk Copper Mines)

Geo-statistical modeling requires a high level of experience and precision that was formerly hand-made and nowadays is semi-computer or computer generated. Most of these models are based on interval data obtained from exploratory drilling cores and linear interpolation techniques. Nowadays, geo-statistical methods, especially indicator kriging, have been found to be highly capable of producing geometric-geological models but most of the time the changes in the geological micro-zones are gradual, and these models have various ambiguities, which is particularly important in the separation of alteration zones in porphyry copper mines.

 Among the data provided during geochemical and geotechnical exploration drilling, the images of drill core boxes are important and comparatively inexpensive. In this study, the quantitative data of red, green and blue were separately used by processing these types of images and converting them into color interval data in the RGB system instead of using the qualitative geological log data. The operation was carried out in a C-Shop program called Core Processing and executed on 20 boreholes of Kahang copper mine. Then, a database of joint density and RGB color system with a lag of 2 meters was prepared. One of the main features of color data is the normality of their distribution function in this region.

This geostatistical study also yields the validated spherical variograms. Initially, a model for the Kahang Mine with a dimension of 700 × 300×400 m was developed and the values of R, G and B for rock blocks of 5 × 5 × 5 were estimated using ordinary kriging. This made it possible to obtain a real colors of geological section based on the triple RGB color system. For the verification of the validity of the proposed method, the color sections were compared with the lithology and alteration sections (already prepared by geologists) of Kahang Mine. For this purpose, a new index was defined as Irgb, which interpreting the length of the vector on the three axes R, G, and B (which did not show good agreement with the lithological sections). However, an acceptable similarity was obtained by comparing the alteration map in 1700 horizon with the Irgb index at the same level in Kahang Mine.

   To ensure the feasibility of using the geostatistical model of RGB color system and the Irgb index another comparison was made at the Miduk Copper Mine and the colored model of this mine was prepared too. The tomography operation was performed in several profiles to determine the mine's hydro zones where the sections of electrical resistance coefficients were already prepared. These sections were then compared with the Irgb index sections prepared from the RGB model proposed in this study.  The comparison was able to detect and distinguish alteration zones due to the sensitivity of the color to the alteration and also to prevent zones with low electrical resistance from being considered as hydro zones by mistake.