Efficient Measurement of Key-Cap Flatness for Computer Keyboards With a Multi-Line Structured Light Imaging Approach

Abstract

A computer keyboard is an important human–machine interaction device. A key-cap and a keyboard are flexibly connected, and its assembly quality, especially flatness, is crucial. In manufacturing quality control, the requirement for assembly quality of keyboards is quite strict, where the flatness measurement of key-caps is a key link. In this paper, an apparatus based on multi-line structured light imaging, and machine vision is designed for online rapid detection of key-cap flatness, is proposed. First, a novel 3D measurement scheme is presented to replace the complete 3D surface reconstruction of conventional methods. Then, an array of cameras with a perspective projection transformation matrix calibration approach is adopted to expand the scanning range of the 3D sensor for large-scale and high-resolution target imaging. In addition, a multi-step image processing is applied to accurately extract the centerline of the projected fringes on a key-cap with printed character, which indirectly helps to improve the measurement accuracy. Finally, a sparse sampling flatness estimation algorithm is proposed to improve the detection efficiency. According to experiments in real production lines, our apparatus can achieve a detection accuracy of 99.74% for various computer keyboards.