Abstract: In the process of developing and utilizing mineral resources, the impact of mining activities in mining areas on the surface environment is increasingly receiving attention. Traditional methods for extracting mining land occupation information exhibit certain limitations in terms of accuracy and reliability. To improve the precision and efficiency of information extraction, this paper proposes a method for extracting mining land occupation information by integrating optimized features and an optimal segmentation scale. This method uses high-resolution GF-2 satellite data as the data source, calculates the optimal segmentation parameters for non-mining land occupation information through the Euclidean II index method, and combines image segmentation technology to determine the optimal segmentation scale for mining land occupation and bare land. In terms of feature selection, this study comprehensively considers spectral, textural, and shape features. By selecting feature factors based on their importance, the performance of the classification model is enhanced. The selected features are associated with the Random Forest （RF） algorithm to construct a classification model with the optimal parameters for extracting surface information within the study area. The experimental results indicate that the method set out in the present study has achieved significant results in extracting surface information from mining areas. The overall accuracy reaches 86.00%, the Kappa coefficient is 0.78, which is an improvement compared to the CART model. This suggests that the combination of Euclidean II index method and image segmentation tools can effectively obtain the optimal segmentation parameters of mining area features, thereby improving the accuracy of information extraction. This method is not only applicable to the primary remote sensing interpretation of mining land occupation, but also helps to quickly and accurately identify the main mining land occupation, improve the accuracy and efficiency of interpretation. In summary, the method presented in this study offers a new approach and technical means for extracting mining land occupation information, which holds significant importance for environmental protection and management in the process of mineral resource development.