In this work, a survey of the diaclases affecting the Sierra de Lihué Calel was carried out with the consequent description, parameterization and classification of the different systems of diaclases in the field, with the aim of determining the origin of this structuring. In addition, the scope and limitations of artificial intelligence techniques based on Machine Learning and Deep Learning algorithms for the detection of joints in the processing of satellite images and photographs of outcrops were evaluated. By means of this method, the generation of diaclase maps and statistical analysis of the length, distribution and orientation of the diaclases in an automated way was carried out.The application of artificial intelligence techniques was compared with traditional methods of analysis, and the accuracy and speed of this innovative methodology was determined, although the chronological analysis and the persistence of joints is limited. Based on the results obtained, modifications to the algorithms were introduced to improve their versatility and the limitations of their application were postulated.