In the implementation of mine development and mine remote sensing environment monitoring, the statistical calculation of monitoring data indicators is one of the main methods for the centralized display and application of monitoring results, and is also the basis for relevant law enforcement and provides data support for the adjustment and formulation of relevant mine policies. In calculating various monitoring indicators, traditional manual statistics have many shortcomings such as low efficiency, high labor cost, large workload, wrong calculation, low accuracy, and inconsistent statistical standards. Through the project practice, the paper puts forward the automatic statistical method of mine remote sensing monitoring based on data stream filtering and multi-dimensional matrix calculation, which can quickly count the attribute information and spatial information of each mine remote sensing monitoring data, realize the calculation of each classification index according to the administrative region and mine type statistics, and complete the rapid summary of mine statistical indicators. The experimental results show that the proposed method completed the statistical summary task quickly, accurately and automatically in the remote sensing monitoring project of Qinghai Province in 2018, thus saving a lot of human resources and time and having a good application effect.