Research on Multi-View 3D Reconstruction Technology Based on SFM

Gao, Lei; Zhao, Yingbao; Han, Jingchang; Liu, Huixian

doi:10.3390/s22124366

Cited by 17 publications

(7 citation statements)

References 27 publications

(28 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, the FLANN matching algorithm [25] matches the feature points to different images. Finally, the traditional 3D reconstruction algorithm -SFM algorithm [26] is used to reconstruct the above-processed photos, which can be divided into incremental, global, hybrid, hierarchical, and semantic-based methods of the SFM algorithm [30][31][32]. A hierarchical algorithm is used for image 3D reconstruction, as shown in Figure 6.…”

Section: Three-dimensional Reconstructionmentioning

confidence: 99%

Brain MRA 3D Skeleton Extraction Based on Normal Plane Centroid Algorithm

Feng,

Zhu,

2023

EAI Endorsed Trans Perv Health Tech

View full text Add to dashboard Cite

INTRODUCTION: Analysis of magnetic resonance angiography image data is crucial for early detection and prevention of stroke patients. Extracting the 3D Skeleton of cerebral vessels is the focus and difficulty of analysis. OBJECTIVES: The objective is to remove other tissue components from the vascular tissue portion of the image with minimal loss by reading MRA image data and performing processing processes such as grayscale normalization, interpolation, breakpoint detection and repair, and image segmentation to facilitate 3D reconstruction of cerebral blood vessels and the reconstructed vascular tissues make extraction of the Skeleton easier. METHODS: Considering that most of the existing techniques for extracting the 3D vascular Skeleton are corrosion algorithms, machine learning algorithms require high hardware resources, a large number of learning and test cases, and the accuracy needs to be confirmed, an average plane center of mass computation method is proposed, which improves the average plane algorithm by combining the standard plane algorithm and the center of mass algorithm. RESULTS: Intersection points and skeleton breakpoints on the Skeleton are selected as critical points and manually labeled for experimental verification, and the algorithm has higher efficiency and accuracy than other algorithms in directly extracting the 3D Skeleton of blood vessels. CONCLUSION: The method has low hardware requirements, accurate and reliable image data, can be automatically modeled and calculated by Python program, and meets the needs of clinical applications under information technology conditions.

show abstract

Section: Three-dimensional Reconstructionmentioning

confidence: 99%

Brain MRA 3D Skeleton Extraction Based on Normal Plane Centroid Algorithm

Feng,

Zhu,

2023

EAI Endorsed Trans Perv Health Tech

View full text Add to dashboard Cite

show abstract

“…These methods have found wide applications in precision agriculture, geographic surveying, archaeology, and others [17]. SfM is a commonly used sparse reconstruction algorithm that aims to reconstruct the 3D coordinates of image feature points [18]. MVS leverages the camera model based on SfM to compute the camera poses and conducts dense reconstruction to recover the geometric information of the scene [19].…”

Section: Introductionmentioning

confidence: 99%

SA-Pmnet: Utilizing Close-Range Photogrammetry Combined with Image Enhancement and Self-Attention Mechanisms for 3D Reconstruction of Forests

Yan,

Chai,

Han

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Efficient and precise forest surveys are crucial for in-depth understanding of the present state of forest resources and conducting scientific forest management. Close-range photogrammetry (CRP) technology enables the convenient and fast collection of highly overlapping sequential images, facilitating the reconstruction of 3D models of forest scenes, which significantly improves the efficiency of forest surveys and holds great potential for forestry visualization management. However, in practical forestry applications, CRP technology still presents challenges, such as low image quality and low reconstruction rates when dealing with complex undergrowth vegetation or forest terrain scenes. In this study, we utilized an iPad Pro device equipped with high-resolution cameras to collect sequential images of four plots in Gaofeng Forest Farm in Guangxi and Genhe Nature Reserve in Inner Mongolia, China. First, we compared the image enhancement effects of two algorithms: histogram equalization (HE) and median–Gaussian filtering (MG). Then, we proposed a deep learning network model called SA-Pmnet based on self-attention mechanisms for 3D reconstruction of forest scenes. The performance of the SA-Pmnet model was compared with that of the traditional SfM+MVS algorithm and the Patchmatchnet network model. The results show that histogram equalization significantly increases the number of matched feature points in the images and improves the uneven distribution of lighting. The deep learning networks demonstrate better performance in complex environmental forest scenes. The SA-Pmnet network, which employs self-attention mechanisms, improves the 3D reconstruction rate in the four plots to 94%, 92%, 94%, and 96% by capturing more details and achieves higher extraction accuracy of diameter at breast height (DBH) with values of 91.8%, 94.1%, 94.7%, and 91.2% respectively. These findings demonstrate the potential of combining of the image enhancement algorithm with deep learning models based on self-attention mechanisms for 3D reconstruction of forests, providing effective support for forest resource surveys and visualization management.

show abstract

“…Then under the framework of structure from motion (SfM) [ 3 ] the spatial information of the scenes can be reconstructed from the images captured over time. For example, Gao et al designed a set of multi-view 3D reconstruction technology based on SFM and obtained the external parameters of the camera and the sparse point-cloud model [ 4 ]. Wang et al integrated the fringe light projection with SfM to realize a global dense point cloud reconstruction system [ 5 ].…”

Section: Introductionmentioning

confidence: 99%

A Virtual Multi-Ocular 3D Reconstruction System Using a Galvanometer Scanner and a Camera

Han

Zhang

2023

Sensors

View full text Add to dashboard Cite

A novel visual 3D reconstruction system, composed of a two-axis galvanometer scanner, a camera with a lens, and a set of control units, is introduced in this paper. By changing the mirror angles of the galvanometer scanner fixed in front of the camera, the boresight of the camera can be quickly adjusted. With the variable boresight, the camera can serve as a virtual multi-ocular system (VMOS), which captures the object at different perspectives. The working mechanism with a definite physical meaning is presented. A simple and efficient method for calibrating the intrinsic and extrinsic parameters of the VMOS is presented. The applicability of the proposed system for 3D reconstruction is investigated. Owing to the multiple virtual poses of the camera, the VMOS can provide stronger constraints in the object pose estimation than an ordinary perspective camera does. The experimental results demonstrate that the proposed VMOS is able to achieve 3D reconstruction performance competitive with that of a conventional stereovision system with a much more concise hardware configuration.

show abstract

Research on Multi-View 3D Reconstruction Technology Based on SFM

Cited by 17 publications

References 27 publications

Brain MRA 3D Skeleton Extraction Based on Normal Plane Centroid Algorithm

Brain MRA 3D Skeleton Extraction Based on Normal Plane Centroid Algorithm

SA-Pmnet: Utilizing Close-Range Photogrammetry Combined with Image Enhancement and Self-Attention Mechanisms for 3D Reconstruction of Forests

A Virtual Multi-Ocular 3D Reconstruction System Using a Galvanometer Scanner and a Camera

Contact Info

Product

Resources

About