| Digital holography is a kind of holographic technology. It has advantages of safeness, non-contact, high resolution and fast imaging in measurement. Therefore, it has been widely used for 3D shape measurement field.The principle of digital holography to reconstruct 3D shape mainly depends on the optical interference and diffraction. Then, the amplitude and phase information are extracted from object wave. According to the relationship between phase and the height of object surface to reconstruct 3D shape. However, the extracted object wave’s phase is mostly wrapped. In general, to achieve the real phase distribution, phase unwrapping is always used. But it exists some deficiencies in practical application. For example, it is sensitive to laser noise and digital noise, and the whole calculation process is complicate and time-consuming. Especially, when the object surface is discontinuous, phase unwrapping can not be realized, even causing large measurement errors.Stereo matching is a technology to acquire 3D scene structure information. Observing the same scene from two perspectives, it gets different views by using camera. Then, the disparity can be achieved from different views through using stereo matching algorithm. At last, obtaining the 3D object scene on the basis of the geometric relationship between the disparity and depth.Aiming at phase unwrapping technology exists some disadvantages in 3D shape reconstruction. This paper presents a method that stereo matching technology used to 3D measurement in digital holography. Digital hologram records the 3D information of the object, and different areas of the hologram can reconstruct different perspective reconstructed images. Firstly, the single digital hologram is divided into two parts. Secondly, two reconstructed images are obtained after the reconstruction, and two reconstructed images are equal to "right view" and "left view" in the binocular vision. However, the two views exist disparity. The disparity can be achieved from different views by using stereo matching algorithm. At last, according to the geometric relationship between the disparity and depth to obtain 3D shape. The validity of this method is proved by numerical simulation and experiment.This paper is arranged as follows:Chapter one, firstly the development of digital holography is summarized, then the study status quo of 3D reconstruction in digital holography is introduced, the application, currently some study issues of digital holography are analyzed and discussed. Computer stereo vision theory is introduced. Finally, proposing the research contents of this paper.Chapter two mainly introduces the basic theory of recording and reconstruction of digital holography. We talk about the relationship between image point of the reconstructed plane and the actual image point, and phase unwrapping theory and problems in digital holography.Chapter three illustrates theory, development status and applications about 3D reconstruction in digital holography based on stereo matching algorithm. We detail analysis region matching algorithm based on block matching technology and graph cut matching algorithm based on mean shift.Chapter four based on 3D reconstruction principle of reconstructed image in digital holography, the relation with the reconstructed images of digital hologram divisions is discussed, the relationship between disparity and depth of reconstructed disparity image is analyzed. Then, we analysis the effect on the factors of reconstructed image depth resolution. It provides a theoretical basis for 3D reconstruction combine with stereo matching and digital holography. Finally, giving computer simulation and experimental results. We test and verify the discrete objects and continuous objects respectively, and obtaining a satisfactory result. The results show that the proposed method is feasible and effective.Chapter five proposes summary and outlook, we sum up the work and point out deficiencies in this paper. |
PDF Full Text Request