Method for Measuring the External Dimensions of Prefabricated Components in Complex Backgrounds

Due to the low efficiency of manual measurement of prefabricated components, there may be defects in the edges and corners of the components, and the images captured in natural environments have complex backgrounds. Based on the theory of binocular stereo vision, this paper proposes a method for measuring the size of prefabricated components in complex backgrounds. Firstly, the principle of binocular stereo vision system is introduced, and the Bouguet algorithm is used to perform stereo rectification on the binocular stereo vision system. Then, the Single Shot MultiBox Detector (SSD) object detection algorithm is used to obtain the position of the object to be measured in the image, filter out the background, and detect key feature points of the object within the detection box. Finally, based on the theory of epipolar constraint, the key feature points of the component are matched, and the key dimensions of the prefabricated component are calculated. Experimental verification was conducted by capturing images in a prefabricated component manufacturing plant, and the results showed that the measurement error of this method was around ± 2%.