To address the issue of multiple calibrations in current printed circuit board (PCB) drilling positioning methods, a novel PCB drilling positioning method that integrated visual calibration and spot detection algorithms was proposed, which could adapt to dynamically changing shooting environments, thereby improving the robustness of PCB drilling positioning. The proposed method was based on Zhang's calibration method, and the camera parameters as well as distortion coefficients were obtained by shooting the calibration board from multiple angles. These parameters were then used to correct the PCB image, and a multi-scale spatial spot detection algorithm was employed to extract the center coordinates of the drilling points in the PCB image. Finally, with the help of camera parameters and image coordinates of PCB drilling points, the three-dimensional spatial coordinates of PCB drilling points were successfully reconstructed, and efficient drilling point positioning was realized. Through the experiment, the positioning accuracy of this method fluctuates around 0.1 mm, with an average positioning error controlled within 0.05 mm~0.07 mm. The results show that this method possesses a certain degree of operational flexibility and robustness, which meets the requirements of PCB mechanical drilling processing.