Method for defects detection and 3D reconstruction based on dispersed points cloud

Equipment surface defect detection is very important for ensuring safety production and avoiding economic losses. Aiming at the algorithm complexity problem of defect detection and 3D reconstruction for dispersed 3D points on the equipment surface, a new method was proposed. All the points were divided into different layers along to the axis, then the points on the layer were moved to the middle plane of the layer. The distribution of the points became regularly and the defect detection and 3D reconstruction were done by the points regularized. The dispersed points defects detection and 3D reconstructions were fulfilled and the error of the defection detection is 1.01%. Experimental results show that this defection detection method proposed is simple and easy to implement after the dispersed points are regularized.