The depletion of natural resources implies the need for a constant search for new reserves to satisfy demand. In the mining sector, Unmanned Aerial Vehicles (UAVs) have revolutionised geo-information capture and modelling to allow the use of low-cost sensors for prospecting and exploration for potentially exploitable resources. A very powerful alternative for managing the huge volume of data is the Geographic Information System (GIS), which allows storage, visualisation, analysis, processing and map creation. The research in this paper validates a new quasi-automatic identification of mining resources using GIS thermal-image analysis obtained from UAVs and low-cost sensors. It was tested in a case that differentiated limestone from dolostone with varying iron content, and different thermal behaviour from solar radiation, thereby ensuring that the thermal image recorded these differences. The objective is to discriminate differences in an image in a quasi-automatic way using GIS tools and ultimately to determine outcrops that could contain mineralisation. The comparison between the proposed method with traditional precision alternatives offered differences of only 4.57%, a very small deviation at this early stage of exploration. Hence, it can be considered very suitable.