As one of the core components of hydraulic system, the surface size of oil distribution plate is directly related to the performance of the hydraulic system. Aiming at the problems of difficulty in extracting partial surface contour features and large measurement error, this paper proposes a spherical segmentation and size measurement method based on additional image edge detection. Firstly, the Sobel operator is adopted to sharpen the edge of the oil distribution plate image, and the contour of the spherical part is extracted to form the mask. Then, the four-step phase shift diagram of the spherical part is segmented by mask, the four-step phase shift method is adopted to solve the phase principal value, and the dual-frequency heterodyne method is adopted to unfold the phase. Finally, the three-dimensional point cloud data of the spherical part of the oil distribution plate is calculated and spherical fitting is carried out through the system calibration. Experimental results show that the spherical partial point cloud image can be segmented clearly by using this method. The measurement accuracy based on point cloud data fitting is 99.238%, which is higher than the measurement accuracy based on height segmentation, RANSAC (Random Sample Consensus) segmentation, region growing segmentation and manual box selection method, and the high-precision measurement of spherical size of oil distribution plate is realized.