Microwave radiometers are indispensable instruments in global environmental monitoring. However, acquiring high-spatial-resolution microwave radiometer imagery faces significant challenges due to complex imaging environments and transmission bandwidth constraints. The core objective of enhancing microwave radiometer image spatial resolution is to reconstruct an ideal apparent temperature field from observed antenna temperature data. Existing resolution enhancement algorithms, such as the BG algorithm, Wiener filter deconvolution, and Lucy-Richardson algorithm, often introduce noise and ringing effects while improving spatial resolution, thereby compromising data quality to some extent. To address this, this paper proposes a resolution enhancement method that integrates edge preservation and denoising capabilities, building upon classical spatial resolution enhancement algorithms. This method innovatively incorporates a bilateral filter to effectively suppress noise in flat image regions. Simultaneously, it cleverly integrates high-frequency gradient information to prevent excessive blurring of sharp image boundaries by the filter. Experimental results demonstrate that this approach significantly enhances the fidelity of microwave radiometer data while achieving spatial resolution levels close to the target.