Abstract: In view of the common industry problems of insufficient identification accuracy of coal seam seismic reflection information and difficulty in meeting the requirements of precise reservoir strata prediction under low signal-to-noise ratio conditions, research on amplitude recovery technology based on surface consistency processing is carried out. Selecting the Maojiachuan measurement area with severe surface undulations, shallow layer low-speed zone development, and complex interference types as the research target area, a multi-level processing flow is designed and implemented for refraction static correction, adaptive surface wave suppression, spherical surface diffusion compensation, and surface consistency amplitude normalization. Through strategies of unified reference surface correction, multi-domain interference wave suppression, and amplitude equalization, the influences of surface conditions, propagation paths, and absorption attenuation on seismic wave energy are systematically eliminated. The experimental results indicate that after processing, the signal-to-noise ratio of the seismic data is averagely improved by 10-15 dB, the horizontal continuity and amplitude stability of the coal seam reflection interfaces are obviously improved, and the fidelity of the seismic data is effectively improved, providing a reliable basic data support for subsequent structural fine interpretation, reservoir strata distribution prediction, and resource evaluation.