Robust and High-Precision Vision System for Deflection Measurement of Crane Girder With Camera Shake Reduction

Crane girder is subject to alternating load in long time, which makes it prone to fatigue and fracture. Vision deflection measurement has become an effective method for structural health monitoring. However, crane girder does not always exhibit high contrast feature and artificial markers have to be pasted. Existing artificial markers cannot perform well in poor light conditions and the accuracy is limited. In this paper, we design an anti-interference marker by pasting a film onto a circular surface light source. The marker allows the vision system to work well under different light conditions. In addition, the reduction of camera shake and sub-pixel technique have been studied to improve the accuracy. According to the characteristics of the marker, corresponding image processing algorithms are also studied. Firstly, the designed marker can suppress environmental interference even at long distances with a wireless module and a small aperture. Secondly, the center of the marker in the image can be calculated using a sub-pixel detection technique. Finally, a camera shake reduction method is applied by using multiple reference markers and scale conversion factors are taken into account to get the physical displacement. To verify the measurement results, two experiments were carried out using a digital dial gauge sensor and laser displacement sensor for comparison. The experimental results proved the robustness and high precision of the vision system.