Error compensation model for monocular vision measurement

Abstract

A monocular vision measurement system was designed in an industry application. In order to reduce the monocular vision measurement error caused by target's deviation from the camera optical axis, camera lens distortion and the platform not ideally vertical to the camera optical axis, an error compensation model was proposed to improve the measurement precision. Experimental results showed that the measurement error was significantly reduced, and the measurement accuracy can reach to within 1 mm even if the distance between the measured target and the camera is about 3 meter.