With an aim to overcome the problem of excessive data transmission bandwidth in near-eye display systems, this study proposes a multiscale pyramid model to realize multi-resolution display in the human visual region. We apply the characteristics of the spatial change rate of the human visual region to the data transmission process. Based on a multi-scale pyramid model, a microdisplay controller for the near-eye display system is designed, realizing the encoding and decoding processes of the video image. The feasibility of this controller is verified on a 1600×1600×3 full-color high-definition OLED-on-silicon microdisplay. Experimental results show that the multi-resolution image fusion algorithm based on a multi-scale pyramid meets the visual perception requirements of human eyes and solves the boundary line effect appearing in the subsampling process, effectively reducing the data transmission bandwidth. The proposed algorithm satisfies the requirements for the real-time rendering performance of the microdisplay in the near-eye display field.