Research On The Key Technologies Of Dynamic Interferometer With Common Paths

The rapid development of modern optics manufacturing industry is closely related to the continuously improvement of its relative detection techniques. With the development of super microfabrication in integrated circuit and aerospaceareas, the requirements on the optics manufacturing and detection techniques is being increased and increased. Thus the precision is on nano-level. Currently, to detect the precision optical elements and the optical systems, the most effective way on ultra-precision detection is the interferometer.Dynamic interferometer of co-optical paths is an opticalstructure widely applied in inference measurement. The advantage is the adoption of common optical path, i.e.the detection arm and standard arm coincide, which greatly improves the detection precision and the application range. This paper presents the system design principles and key technologies of a new type of Dynamic interferometer of co-optical paths.Through the adoption of polarized light inference, it can obtain four interferogramswith their phases having a π/2-difference in sequence at the same time.Itrealises the high precision detection meanwhile it reduces the influence from the air flow and the environment vibration, which is adapt to the detection of wide aperture and long focus optical elements. This interferometer can be applied to the field inspection of optics system, optical inspection in complex and harsh environment,and constant dynamic detection, i.e. the detection of the change process on opticalsurface.Through studying and analysing the factors that influence the optical instruments detection, this paper proposes a practical opti mechanics structure of interferometer, especially the structure design of theoptical registration. The research methodology of this paper is: based on the traditions interferometer principles, we design thesynchronised receiving optical path for four-phase interferometer, which is a breakthrough to the implement principles; we study the synchronised acquisition techniques of CCD co-phase surface of four-phase detection; we also study the de-noising algorithms of interference fringes and the algorithms of transforming between fringes information andsurface information; by analysing the system deviations of the dynamic interferometer of co-optical pathsand the influence from the random deviations, we propose an algorithm to erase the deviations.In theopti mechanics structuredesignofthe dynamic interferometer of co-optical paths, we take use of rotary structure of thepolarised splitting prism to shift the space phases of the four adjacent inference fringes which have 90°phase difference. To solve the optical registration problem during the synchronised acquisition process,we design the registration instrument for the inference fringes acquisition. By the optical and mechanical structure design, the physical registration for the fringes images acquisition is realised. Based on this, we take use the digital image processing methods for high precision image registration, which increases the detection precision.Considering the data process method, to solve the noise problem in the acquired fringe images, we propose anelectronic speckle fringe filtering method based on theanalyses of wavelet within the isoline window. Compared with the classical median filtering andthemean filtering, our method decreases the influence on the unwrapping algorithms meanwhile has a better signal-to-noise ratio.To reduce the influence on the system precision form the system deviations and the random deviations, we propose a rotary deviationcorrection model to adjust the rotary asymmetric deviations during the optical surface absolute detection. Firstly,based on the theory of the classical N-step rotation average method, the mathematical expression of surface deviation is presented via Zernike polynomials.Then the Zernikecoefficients are adjusted according to the deviation caused by the rotation anglesand as a result the rotary asymmetric deviations are corrected. On the promises of the correctness of the correction model verified by the numerical simulation methods, the model will be used to process the practical detection data.