Correction methods of laser intensity and accuracy of point cloud classification

The linear, logarithmic and cubic polynomial physical correction models were established based on the laser radar equation to correct a variety of factors relating to laser intensity. Then, the corrected intensity values were used for point cloud classification, and the correction effect, classification accuracy and computing efficiency of the three models were analyzed and compared. The results show that the linear model owns the lowest classification accuracy and the smallest amount of computation, while the cubic polynomial model possesses the highest classification accuracy and the largest amount of computation and the logarithmic model takes accuracy and computation into account at the time. Therefore, one of the three proposed models can be chosen to be applied to data processing according to the amount of the collected data, and the requirements of computation efficiency and classification accuracy. Point cloud after classification can provide data base and feature constraints for subsequent modeling to achieve the goals of effective compression and efficient point cloud modeling.