This paper proposes a novel Range Migration Algorithm （RMA） integrated with an adaptive background filtering method specifically designed for near-field millimeter-wave imaging scenarios where targets are in close proximity to background structures. This method simulates the attention distribution mode of the human visual system which is used in Artificial Intelligence （AI） and called the Attention Mechanism. Based on the concept of static clutter filtering， the frequency-domain signals of the scanning aperture are divided into grid cells. Background scattering functions are established by analyzing the motion processes within each cell， and the background interference is linearly filtered out. An analysis of the manifestation of background scattering interference within the algorithm is carried out， and the impact of the grid cell dimension on the imaging quality is investigated. Experimental results show that the proposed method exhibits the capability to enhance the signal-to-noise ratio of both the target and the background. It effectively suppresses the background interference， leading to a more prominent image， meanwhile without imposing the excessive computational load. The method offers a novel solution for improving the performance of millimeter-wave imaging technology in practical applications.