Aimed at the adjustment and optimization of enterprise productivity structure, a JBZ23.07 rotortype oil pump intelligent assembly system based on machine vision was proposed. According to the requirements, the overall scheme was formulated, from the whole to the part’ was used as a design method, the intelligent assembly system was decomposed into visual system, upper computer software system and other parts for research. Halcon software was used to complete the camera calibration and distortion correction; the shapebased multitemplate matching method was used to extract the relevant features of the part image, and the pose adjustment parameters were obtained. Through many tests, it can be seen that the image processing time of a single part is 130 ms and the position angle error is 0.01°. The accuracy, stability and sensitivity of the algorithm were verified by multimethod and multiangle experiments. The results show that the average identification success rate of the algorithm is 97.60%, 100.00% and 98.04%, respectively, which meet the overall design requirements and actual production requirements.