Noise reduction method for lidar echo data based on multivariate analysis method

A new noise reduction method for lidar echo data is proposed. This method is based on canonical correlation analysis between two multivariate data groups. GROUP-I dataset includes lidar echo data with lasing and without lasing. GROUP-2 dataset includes moving averaged data of GROUP-1. As the experimental results using both artificial data and actual lidar echo data, the amplitude of noise component in the primary canonical variate of GROUP-1 (result) is reduced more than 30%.<<ETX>>