In aero-engine manufacturing, accurately acquiring the profiles of complex aviation components using advanced measurement techniques is essential for machining guidance and quality assurance. To improve the measurement performance of structured-light systems on complex structural surfaces, this article proposes a multi-resolution structured-light measurement method. A high-resolution main camera and a low-resolution auxiliary camera are used in a binocular structured-light system, and the phase-matching strategy is optimized to enhance measurement quality. First, the resolution of the main camera is increased to capture fine surface features, and high-precision calibration of the stereo vision system with cameras of different resolutions is achieved based on photogrammetry. Then, an epipolar-constrained phase-matching method is adopted. By precomputing the disparity range between the left and right views within the measurement field, the search range for phase-corresponding points along the epipolar line is constrained, thereby improving matching efficiency and reducing redundant computation. Finally, the edge type of each phase-corresponding point is identified by calculating the neighborhood gradient of its initial value, and edge interpolation is optimized to improve edge-measurement completeness. Experiments show that the calibration reprojection error is less than 0.15 pixels, the surface measurement accuracy for complex aero-engine components reaches 5 μm, and the edge-feature measurement accuracy reaches 12.1 μm. The disparity-constrained epipolar search improves the search speed of phase-corresponding points by 88.1%. Compared with the dual-low-resolution structured-light scheme, the proposed method achieves better measurement performance, approaches the dual-high-resolution system, and overcomes the limitation of constructing stereo vision systems with cameras of different resolutions.