Study On Large Aperture Optical Plane Detection Algorithm Based On Image Processing

With the increase in optical applications demand, the corresponding optical detection requirements also have been more and more highly, the main changes are the request of precision and diameter of component detected. In some applications, large aperture optical plane is a key component of the whole system, which will directly determines the systematic view and space resolutions. The size of the optical components, related with optical signal, is gradually expanding, which hasn’t met the requirements of direct detection in the existing test equipment. This project is on large diameter flat surface shape detection. A detection method, based on image processing technology, was proposed in the project, which detection algorithm is the key to research. This test instrument is the small interferometer of Zygo in this test method. According to planning of sub-apertures beforehand, the first step is to collect images for every sub-aperture, and then the related images will be processed with the help of image processing technology, there will be a series of algorithms to complete the image preprocessing, including image pre-processing, image matching, image fusion, image analysis and surface fitting. This research tasks are as follows:First, in this thesis, the principle of sub-aperture interference detection and the existing sub-aperture stitching methods were introduced, By reference to the related martials on sub-aperture partition and the arrangement mode of sub-apertures, the corresponding position and value of sub-apertures were given according to the experience in the researching field.Second, Matlab was used to simulate the phase image based on interference, and then followed the sub aperture stitching scheme by taking the method of the mask, then got a series of overlap sub aperture images covering large aperture.Moreover, in view of the actual situation of surface plane interference detection, according to the following image sequence:preprocessing, registration and fusion, stitching algorithms were analyzed.Finally, the method was verified through the experiment, which was on a diameter of 150 mm optical plane measured by 100 mm Zygo interferometer in the laboratory, then the collected interference images and related data, and then were processed according to the previous algorithms, and the results was analyzed on the detection evaluation indexes to further proves the detection algorithm is feasible.