Research On The Application Of Digital Image Processing Technology In Zeeman Effect Analysis

Zeeman effect refers to the phenomenon that the energy levels of atom split in the external magnetic field. It illustrates that light is also affected by magnetic field. Zeeman effect domonstrates space quantization of magnetic moment, and thus makes it possible to study atomic structure. Besides, the charge-mass ratio of electron can be measured applying Zeeman effect.The distinctive feature of the study of Zeeman effect in recent years is that it has adopted the technology of digital image processing. In this paper, the author discusses first some of the related theories concerning digital image processing, such as image enhancement, binarization and the thinning of binary image. And then he explores feature measuring and extraction for digital image by using Hough Transform.To be specific, this paper focuses on the whole process of digital image processing of splitting spectral lines in Zeeman effect. By using WPS-Zeeman effect instrument, especially by applying the method of turning Polarizer combined with CCD image acquisition system, get theπcomponent's are spectral line interference image of Zeeman splitting. Because of the distributive features of spatial light intensity in interference, the fringe has a certain width. By applying the technologies of image preprocessing, gray level transformation, weighted smoothing, Yatagai extreme determination and image thinning, we can thin the fringes with certain width in the splitting spectrum images to binary images with single-pixel width.This paper adopts three-dimensional processing method while extracting information of fringe by way of Hough Transform. Based on prior knowledge of Zeeman effect splitting spectrum, limit the choice range of fringe radius, and thus save great effect in calculation. Through this experiment, get some information about fringes, reconstruct the image and compare it with the previous thinned one, so it is an effective way. In this way, the wave number difference is 56.21m-1, and the Charge-mass Ratio is 1.7645×1011 c·kg-1. Compared with the theoretical value, the percent error of the Charge-mass Ratio is 0.26%. This method is better than the Curve Fitting Method widely used in the present literatures.