Large size and high precision are the main trends in the current development of space antennas. Due to their strong directivity and ease of scanning, parabolic cylindrical antennas are widely used in radar, space power synthesis, remote sensing, and other fields. This paper introduces the main factors affecting antenna surface accuracy, including antenna gain, aperture efficiency, and surface errors. Considering the specificity of antenna configurations and the demand for high-precision automated detection, this paper proposes the HR-ICP point cloud registration method to detect surface errors of parabolic cylindrical antennas. Finally, through verification using simulation data and actual point cloud data, the surface error was less than 3 mm, and the focal length error was less than 1.5 mm. The results demonstrate that this method can effectively detect surface errors of cylindrical antennas and achieve the required precision for design.