Screw pitch precision measurement using simple linear regression and image analysis

In this paper, a novel and simple method of calculating the screw pitch from a screw image is proposed. We combine following three techniques to synthesize the system: (A) calibrating the camera parameters based on the principles of simple linear regression and geometry optics, (B) hiring image processing techniques to extract the screw object from an image, (C) using Hotelling transform to rotate the principle axes of the extracted screw object image to coincide with the coordinate axes. To explore the utility and demonstrate the efficiency of the proposed scheme, simulations under various conditions are conducted. The experiment results show that our proposed scheme can exactly measure the pitches of screws with an error of less than 0.1 μm. This precision is enough and suitable exactly in screw applications.