Method for measuring geometric dimensions of micro parts based on a telecentric camera.

With the development of digital precision manufacturing technology, the demand for precision measurement is continuously increasing in fields such as semiconductors, electronic circuits, and 3C. The measurement method based on vision is widely used in precision measurement. However, an ordinary optical lens has large distortion and variable magnification, which limits the improvement of measurement accuracy. Based on this problem, a method for measuring geometric dimensions of micro parts based on a telecentric camera is proposed, where precise calibration of the telecentric camera and detection of sub-pixel edges are the key factors to determine the measurement accuracy. So, a high-precision calibration method of the telecentric camera based on a single image is proposed, which ensures the calibration accuracy and improves the calibration efficiency. Combining the idea of the Steger algorithm, an improved Zernike moment sub-pixel edge detection method is proposed to improve the detection accuracy of sub-pixel edges. Additionally, a measurement system was established to conduct the experiments of camera calibration, system calibration, and practical performance evaluation to verify the proposed method. The results demonstrated that the relative error of the magnification of a telecentric lens is 0.7%, and the maximum mean absolute error of the measurement result is 0.84μm in the measurement range of 2.1 mm. The proposed method in this paper has high precision and good stability, which can provide technical support for the inspection of the manufacturing precision of micro parts.