KEYWORDS: Point clouds, 3D image processing, Image segmentation, 3D modeling, Denoising, Image restoration, 3D metrology, Fringe analysis, Cameras, 3D image reconstruction
In processing the point cloud of three-dimensional (3D) reconstruction, various point cloud noises would generate, resulting in decreasing the accuracy of 3D point cloud. In order to improve the accuracy of 3D point cloud, this paper proposes a point cloud denoising method based on 3D point cloud segmentation. This method can be divided into four steps. Firstly, the K Nearest Neighbor (KNN) algorithm is used to construct a KNN look-up table for 3D point cloud. Secondly, the look-up table is used to segment the 3D point cloud to obtain the noise point cloud and the noise-free point cloud. Then, using the relationship between the noise point cloud and the noise point of the absolute phase, the reference noise-free phase is established to restore the absolute phase of the noise-free point. Finally, 3D reconstruction is performed according to the recovered absolute phase to obtain a noise-free 3D point cloud. This proposed method only calculates the KNN once for the initial 3D point cloud and only uses a KNN look-up table to segment the point cloud, which also accelerates the speed of 3D point cloud segmentation. The experimental results show that this method not only removes the noisy point cloud, but also restores part of the noisy point cloud into a noise-free 3D point cloud, improving the accuracy of the 3D point cloud.
The passive depth map recovery based on light field data always suffers from low accuracy and high computational burden due to the noise in sub-aperture images and the narrow disparity between sub-aperture views. To overcome this problem, in this paper, temporal digital fringe projections are used to guide the accurate depth map recovery with light field camera. Since the temporal structured light field casts a set of high signal-to-noise-ratio projections onto the object, the depth map of sub-aperture image can be recovered pixel-by-pixel independently. Therefore, the accuracy of the depth map recovery could be significantly improved. In our propose method, the sub-aperture images with temporal fringe projections are extracted from structured light field information, wrapped phase maps of sub-aperture images are obtained from phaseshifting fringe analysis, the absolute phase map and depth map are recovered from a group of wrapped phase maps with temporal phase unwrapping. In our experiment, when the distance between the light field camera and reference plane is 36 cm, the measurement error of the proposed method is within 0.2 mm. The experimental results validate the effectiveness of the proposed method.
Phase retrieve is an important step for phase shifting profilometry (PSP). The existing phase retrieve methods can obtain the phase value successfully for static object. However, as multiple fringe patterns are required in PSP, when the object has movement, errors will be introduced. A new phase retrieve method for the object with 2D movement is proposed in this paper. The 2D movement is divided into translation movement and rotation movement. Then their influence on the phase value is analyzed and a new reconstruction model including the movement information is given. At last, the phase value is retrieved based on the new reconstruction model. The proposed method can eliminate the errors caused by 2D movement of object. The effectiveness of the proposed method is verified by simulations.
In fringe projection profilometry, three-dimensional reconstruction result with higher spatial resolution could provide more detailed description of the measured object. To increase the spatial resolution of three-dimensional reconstruction result, this paper proposes a method to improve the resolution of the absolute phase map recovered in fringe projection profilometry with the digital images of different resolutions. In this method, the same fringe patterns are sampled with the images of different resolutions, the absolute phase maps of different resolutions are obtained respectively. Since the sampling points of the digital images under different resolutions are not coincident, additional ground truth depth information of the object surface is obtained. To emerge an absolute phase map with higher resolution, an algorithm is developed to fuse the absolute maps in different spatial resolutions together. The proposed method can be used to obtain a three-dimensional reconstruction result with higher spatial resolution for various applications. The effectiveness of our proposed method is validated by simulation results.
With the increasing integration level of components in modern electronic devices, three-dimensional automated optical inspection has been widely used in the manufacturing process of electronic and communication industries to improve the product quality. In this paper, we develop a three-dimensional inspection and metrology system for semiconductor components with fringe projection profilometry, which is composed of industry camera, telecentric lens and projection module. This system is used to measure the height, flatness, volume, shape, coplanarity for quality checking. To detect the discontinuous parts in the internal surface of semiconductor components, we employ the fringes with multiple spatial frequencies to avoid the measurement ambiguity. The complete three-dimensional information of semiconductor component is obtained by fusing the absolute phase maps from different views. The practical inspection results show that the depth resolution of our system reaches 10 μm . This system can be further embedded for the online inspection of various electronic and communication products.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.