Image-based displacement measurement using image recognition algorithms is a novel digital displacement measurement technology that offers high robustness and fault tolerance. To reduce the volume of the measurement system while achieving high-resolution linear displacement measurement， a linear displacement measurement system based on spherical light projection imaging was designed. Firstly， the magnification effect of spherical light projection was utilized to design the spherical light projection imaging magnification optical path， and the pixel grayscale values captured by the image sensor were compensated based on the optical wave plane model. Subsequently， to achieve absolute position measurement and encoding element discrimination on a single code track， a pseudo-random sequence encoding method and a decoding method based on width threshold were designed. Then， to achieve sub-pixel linear displacement measurement and enhance system accuracy， a linear displacement subdivision algorithm and a multi-centroid fusion algorithm were designed. Finally， an experimental setup was designed to experimentally validate the proposed method.The measurement optical path height of the designed experimental setup is only 5 mm. The experimental results indicate that the proposed method can achieve a theoretical resolution of 5 nm within a 250 mm range， with an error range of -3.3 μm to 3.6 μm， and after compensation， the error range is reduced to -0.6 μm to 1 μm. The research conducted can provide a technical basis for high-resolution linear displacement measurement studies while reducing the system volume and achieving high-resolution linear displacement measurement.