The frequency limitation of 3D seismic data leads to the insufficient recognition accuracy of small-scale geological bodies such as thin interbeds and coal seam bifurcation and merging. For this reason, a Joint Lateral Constraint Compressed Sensing Multi-Trace Sparse Inversion Method (JL-CSMTSI) is proposed. Under the framework of compressed sensing sparse inversion, this method constructs lateral continuity constraints by combining TV regularization terms with geological structure constraints, enabling collaborative inversion of multi-trace reflection coefficients. This approach effectively broadens the frequency band of seismic data and enhances seismic resolution. In terms of algorithm implementation, an adaptive parameter selection strategy is proposed to overcome the limitations of empirical assignment. To objectively evaluate resolution improvement, quantitative metrics—Effective Bandwidth Ratio (EBR) and Wavelet Compression Ratio (WCR)—are established. Tests on wedge models and the Marmousi2 model demonstrate the method's effectiveness in high-resolution processing of seismic data with complex structures such as thin interbeds, steep dips, and faults. In practical seismic data processing, the method not only preserves the quality of the original data but also improves the identification accuracy of thin interbeds and microstructures, exhibiting strong reliability. The results demonstrate that the JL-CSMTSI method provides a high-resolution processing technique for 3D seismic exploration in coal mining areas, with good potential for engineering applications, and is also applicable to high-precision seismic exploration in other complex geological conditions.