3D laser scanner calibration method based on invasive weed optimization and Levenberg-Marquardt algorithm

To estimate the measurement model parameters of a three-dimensional (3D) laser scanner, a spatial sphere-based calibration method by combining invasive weed optimization (IWO) and Levenberg-Marquardt (LM) algorithm is proposed in this paper. A nonlinear objective function of the 3D laser scanner calibration is established according to the distance constraint between each laser scanning point on a spherical calibration object and the center of the calibration sphere. To reduce the influence of measurement outliers, the Pseudo-Huber loss is also employed in the objective function. After that, a hybrid optimization algorithm combining IWO and LM is adopted to optimize the objective function, where the optimal model parameters of 3D laser scanner are sought. Thus, the model parameters calibration of the 3D laser scanner is realized. Experimental results demonstrate the high calibration accuracy and good consistency of model parameters of the proposed method. And the influences of the measurement noise and outliers can also be effectively suppressed.