The fault-fracture reservoirs in Chengdao oilfield contain faults and fractures, with excellent oil sources and high single-well production. Due to their great depths, the poor quality of conventional migration data, and inadequate imaging of the buried hill interiors, there are significant difficulties in accurate reservoir prediction and characterization. To address this challenge, this study proposes a novel method for characterizing buried-hill fault-fracture reservoirs using full-azimuth angle data with the ES360 processing system. The workflow begins by comparing reservoir responses across different reflection and azimuth angles to select optimal seismic gathers. Following flattening along the strong reflection interface at the top, wavenumber-spatial domain filtering is applied to remove the weathering crust and overlying/underlying sedimentary layers, thereby enhancing features related to oil-and gas-filled fractures and cavities within the buried hills. Subsequently, the first eigenvalue is computed from the background-removed data volume, converting reflection event characteristics into three-dimensional structural representations of the fault-fracture system. Finally, the eigenvalue volume is structurally restored to its original stratigraphic configuration via inverse flattening, achieving effective interpretation of the fault-fracture reservoirs. Application to actual seismic profiles and attribute analysis along the target layer demonstrates that the proposed method effectively highlights the characteristics of steeply dipping buried-hill fault-fracture reservoirs. This approach provides a new conceptual framework applicable to other geologically similar regions.