A field measurement method for large objects based on a multi-view stereo vision system

Abstract

Field measurement of large objects is important in many fields of study. In this paper, an effective field measurement method for large objects based on a multi-view stereo vision system is proposed. First, to reconstruct the dimensional features of large objects, partially saturated laser stripes are projected onto the surfaces of large objects, and a light stripe model of saturated laser stripes is proposed to extract the feature points of the stripes. Then, a flexible laser-aided pattern is designed, and a high-precision feature extraction method based on a region-of-interest (ROI) is developed, by which the local dimensions of large objects acquired from different views can be transformed into the same coordinate system to complete the measurement of the object's overall dimensions. Experiments are performed both in a laboratory and a forging workshop; the laboratory experiments validate that measurements can be completed with only a small overlapping region (25%) in a field of 8.6m×5.0m with a precision of up to 0.10% using the method proposed in this paper. The experiments in the forging workshop show that the method is effective at measuring the height of a large hot forging.