The A oilfield in Bohai Bay is a typical braided delta deposit. Due to the deep burial of the fore-accumulation reservoir, the seismic data have a low main frequency and insufficient longitudinal resolution, which affects the accuracy of the characterization of the single-stage fore-accumulation reservoir. To address this issue, a generalized Beta wavelet domain seismic high-resolution processing method based on non-convex sparse constraints is proposed. Firstly, this high-resolution seismic processing method is represented as an inversion problem with regularization constraints in the wavelet domain, and the non-convex sparse constraints are introduced to obtain more robust and accurate attenuation compensation results, thereby enhancing the longitudinal resolution of seismic data. Secondly, the generalized Beta wavelet transform is introduced to obtain a better time-frequency spectrum that more closely matches the seismic data, and then compensate for the attenuation of each frequency. Finally, the validation results of the noise-free and noisy synthetic seismic data and 2D seismic data of A oilfield show that the proposed processing method can effectively improve the longitudinal resolution of seismic data. The relationship between the strong and weak seismic signals is obvious, and the transverse changes of the in-phase axis (pinch-out) are prominent, with the matching results of the seismic-well ties being in good agreement. Utilizing the research results to interpret the fore-accumulation can accurately characterize the lithologic trap boundary, providing technical support for the subsequent well location deployment and well trajectory optimization.