In comparison with traditional photoelectric displacement measurement technologies, displacement measurement methods based on digital image processing exhibit superior fault tolerance and flexibility, making them a current research hotspot. To achieve high-precision and high-reliability angular displacement measurement, an image-based angular displacement measurement system based on Manchester coding was proposed. First, a single code-channel raster code disc was designed using Manchester coding based on M-sequence pseudo-random coding. A digital image sensor was then used to construct an optical path for capturing patterns on the raster code disc. Subsequently, a decoding recognition algorithm tailored to the coded patterns was developed. Additionally, edge positioning and sub-pixel subdivision algorithms for coded marker edge pattern fitting were proposed to further enhance the system’s resolution. The proposed method was then experimentally validated. The experimental results demonstrated that the system achieved a resolving power of 21 bits and an accuracy of 1.73 arcseconds with a 100 mm grating code disc. This research provides a foundation for the development of highly reliable and high-performance photoelectric angular displacement measurement technologies.