Papers by Keyword: 3D Reconstruction

Abstract: At present, in the most of the digital image system, the input image is used to freeze the multi-dimensional image scanning way again into a one dimensional signal, then carries on the processing, storage, transmission and processing. Finally tend to form multi-dimensional image signal, and the image noise will also be decomposed and compounded. Electrical systems and outside influences in these procedures will enable precise analysis of complexity of image noise. According to off-line learning method of neural networks, this paper focus on the noise filter in the 3Dreconstruction process in order to make the image clearer.

Abstract: On the basis of multimedia digital display technology, combined with wireless network three-dimensional global control, we design a computer virtual simulation system of low carbon tourism scenic spot. And we use the wireless network device WG511, WG311, WG111 and PDA to cover the whole low carbon eco tourism scenic spot. In order to test the availability and reliability of the wireless network coverage model, we test the performance of wireless networks, and obtain the three-dimensional distribution map of signal strength and contour of wireless signal. Through 3D capture on the scenic tour key point, we realize the three-dimensional reconstruction of low carbon sustainable ecological landscape, which provides technical support for the design of the scenic tour landscape.

Abstract: Abstract. Currently, traffic law enforcement agencies rely mainly on the license plate to identify illegal vehicles. But for those vehicles with no license plates or license plates are obscured or slipped, the existing techniques do not work. In this paper, a method to reconstruct the external dimensions of vehicles from 2D images is therefore proposed to identify those illegal vehicles above. The external characteristics of vehicles and methods to model vehicles are analyzed. Based on the analysis, we simplified vehicles as a combination of a rectangular parallelepiped and trapezoidal polyhedron which can be produced with corner points from 2D images. The process to establish the vehicle model with corner points is studied. Finally with Visual C++ and SolidWorks API, the 3D model of a vehicle is established based on a 2D image. The vehicle's primary outlines information including length, width and height are restored, and the information can be used to identify the vehicle.

Abstract: The modal of large flexible structure with low and close modes hardly computes accurate data in the ground. While binocular stereo vision is valued for non-contact measurement, greater flexibility and higher precision in on-orbit identification method. This paper optimized the binocular vision algorithm. The difference between original and new algorithm, which includes the aiming-point number, the identification time and the memory number was shown. The conclusion manifests that new algorithm can improve the identification accuracy and speed up the identification and provide support for on-orbit modal identification.

Abstract: This paper puts forward the method of acquiring underwater target’s 3D coordinate. It also illustrates the underwater linear structured light measure system. With ICCD technology reducing the influence of underwater laser scattering and Steger algorithm extracting the center of the laser stripe, it presents the method of acquiring underwater target’s 3D coordinate. In the end the method is proved by experiment.

Abstract: Mainly studying the three-dimensional reconstruction for multiple Han Dynasty stone images, this paper proposes a 3D reconstruction algorithm based on two images. The author analyzes the basic principle of SIFT matching points detection according to Epipolar geometry constraints and projective transformation of images in 2D plane, and establishes mathematical model for the 3D reconstruction on foundation of sequence images. The feasibility of 3D reconstruction based on the sequence images is approved by simulation for three gray-scale Han Dynasty stone digital images which provides a new research way for identifying the feature of a target by the camera.

Abstract: Based on the fuzzy control cluster algorithm, we establish the modern charm mathematical model of traditional toys with UG software, and design the 3D reconstruction system of traditional folk toys through program. In order to verify the effectiveness of the system, we use the UG software to do 3D reconstruction simulation on the modern charm of traditional toys, and obtain the 3D key point capture graph. Through the key point we construct 3D key line of toys. From the statistics of performance, the maximum frequency of key point reconstruction is up to 8500 times /s, which has good performance. Finally, through the UG 3D reconstruction we get the modern fashion reconstruction design for traditional toy, which provides the technical reference for the study of traditional folk toys.

Abstract: This paper takes the following measures to solve the problem of 3D reconstruction. Camera calibration is based on chessboard, taking several different attitude images. Use corner point coordinates by corner detection to process camera calibration. The calibration result is important to be used to correct the distorted image. Next, the left and right images should be matched to find out the object surface points’ imaging position respectively so that the object depth can be calculated by triangulation. According to the inverse process of projection mapping, we can project the object depth and disparity information into 3D space. As a result, we can obtain dense point cloud, which is ready for 3D reconstruction.

Abstract: The Point Cloud Library (PCL) is a good tool for point cloud data processing. In this paper, a method of 3D reconstruction for rock mass based on PCL is introduced, where hardware choosing, parallel computing, PCL, and edge extraction are analyzed and used in order to realize a better reconstruction effect, including both precision and speed. The reconstruction results can be used in engineering calculation.

Abstract: In this paper, we propose an efficient way to produce an initial transposed matrix for two point clouds, which can effectively avoid the drawback of local optimism caused by using standard Iterative Closest Points (ICP)[ algorithm when matching two point clouds. In our approach, the correspondences used to calculate the transposed matrix are confirmed before the point cloud forms. We use the depth images which have been carefully target-segmented to find the boundaries of the shapes that reflect different views of the same target object. Then each contour is affected by curvature scale space (CSS)[ method to find a sequence of characteristic points. After that, our method is applied on these characteristic points to find the most matching pairs of points. Finally, we convert the matched characteristic points to 3D points, and the correspondences are there being confirmed. We can use them to compute an initial transposed matrix to tell the computer which part of the first point cloud should be matched to the second. In this way, we put the two point clouds in a correct initial location, so that the local optimism of ICP and its variations can be excluded.