Here, we propose a solution to enhance the accuracy of optical encoder based on self-imaging of micrograting by integrating signal correcting and noise suppressing techniques. The method employs an arctangent subdivision algorithm for signal correction asssociated with an optimized noise suppression module, which can effectively address the amplitude imbalance, phase shift, direct current (DC) offset and various noise issues. The hardware circuit integrates voltage regulation and low-pass filtering functions, resulting in suppression of power inter-ference and high-frequency noise. The system is implemented on an FPGA platform. The experimental results indicate an enhancing measurement stability (standard deviation) by 66.88%. Further more, by using optical gratings with period of 4 μm, we demonstrate a displacement measuring precision down to 10-nm level and an accuracy of 0.15 μm in experiment. When applied to a 20 μm-grating-period encoder, the interpolation error is measured to be ~ 0.09 μm. By using the proposed method, a noteworthy higher accuracy and stability of self-imaging-based grating encoders can be expected, showing well potential for this method in semiconductor processing and micro-nano manufacturing