Extracting mode components in laser intensity distribution by independent component analysis

With increasingly sophisticated laser applications in industry and science, a reliable method to characterize the intensity distribution of the laser beam has become a more and more important task. However, traditional optic and electronic methods can offer only a laser beam intensity profile but, cannot separate the main mode components in the laser beam intensity distribution. Recently, independent component analysis has been a surging and developing method in which the goal is to find a linear representation of a non-Gaussian data set. Such a linear representation seems to be able to capture the essential structure of a laser beam profile. After assembling image data of a laser spot, we propose a new analytical approach to extract laser beam mode components based on the independent component analysis technique. For noise reduction and laser spot area location, wavelet thresholding, Canny edge detection, and the Hough transform are also used in this method before extracting mode components. Finally, the experimental results show that our approach can separate the principal mode components in a real laser beam efficiently.