Accurately obtaining attribute information of military geological elements can effectively support the construction of the battlefield geological environment guarantee system and provide basic data support for military decision-making. Traditional methods for obtaining military geological elements, such as field surveys or manual interpretation of remote sensing images, are characterized by high costs, low efficiency, and uncertain accuracy when obtaining data from unfamiliar areas. Utilizing known regional data and machine learning and deep learning methods to construct a military geological element attribute model for hyperspectral satellite imagery has become an effective means of obtaining data from unfamiliar regions. This paper proposes a machine learning-supported method for predicting the basic quality level of rocks, which performs well in predicting the basic quality level of rocks in unfamiliar areas. Based on collecting basic quality grade data of rocks in the research area, a sample dataset was created using the Resource One 02E hyperspectral satellite data as the data source. SVC, RF, XGBoost, Stacking, Blending, and ResNet50 machine learning methods were employed, applied to study the prediction model of the basic quality grade of rocks in unfamiliar areas. The research results show that the ResNet50 model is the best prediction model for the basic quality grade of rocks in the study area, with a prediction accuracy of 65.53%. The Stacking model follows with a prediction accuracy of 41.53%, and the blending model has the lowest prediction accuracy. The overall prediction results of the model reflect a clear spatial differentiation of the basic quality grades of rocks in the study area, presenting a spatial distribution characteristic of higher basic quality grades of rocks in the north, mainly grades I and II, and lower basic quality grades of rocks in the southwest, mainly grades IV and V. The basic quality grade of rocks in unfamiliar areas is mainly grade III or below, with higher basic quality grades in the northeast direction. The research aims to provide a basis for the acquisition and application of military geological data.