In off-axis digital holography and non-coherent light imaging research,spatial frequency domain filtering plays a crucial role in reconstructing both the amplitude and phase information of the target object.The selection and precise localization of the signal spectrum have a profound impact on the quality of the phase reconstruction after filtering.Even minute pixel-level differences can lead to severe distortions in the reconstructed phase.To address this challenge,a method for signal spectrum localization and interference suppression in digital holographic microscopy is introduced.The approach adapts the filtering shape and size based on the histogram distribution characteristics of the holographic spectrum.Additionally,it employs the centroid method for rapid and accurate localization of the signal spectrum's center and employs morphological processing in the frequency domain to eliminate interference in the vicinity of the signal spectrum.:This method achieves adaptive filtering of the signal spectrum,thereby improving the quality of phase reconstruction.The phase simulation in this study was conducted using a two-dimensional mountain-shaped graph,and experimental validations were performed by setting up the optical path to capture the digital area on the USAF 1951 resolution test chart as well as onion epidermal cells.Simulation and experimental results demonstrate that this algorithm can achieve high-quality phase reconstruction while effectively suppressing interference,enhancing the imaging quality of off-axis digital holography systems.