With the increasing development of geothermal energy resources, obtaining precise subsurface thermal conductivity structures has become crucial. However, current geophysical inversion methods lack a detailed technique for directly estimating subsurface thermal conductivity, especially when utilizing both borehole temperature field data and surface heat flow data as constraints. To address challenges posed by sparse borehole data and the limited resolution of borehole temperature field data, this study introduces a novel approach. It first utilizes boundary detection techniques to refine the extent of anomalous regions using heat flow data. Subsequently, by incorporating inversion results from borehole temperature data, a reference model is established, enabling a joint inversion technique that leverages both borehole temperature field data and surface heat flow data. Model experiments demonstrate the feasibility and effectiveness of this joint inversion method, significantly improving subsurface thermal conductivity imaging. Finally, the analysis of field data further validates the practicality and efficiency of this approach.