WiMi Hologram Cloud Inc (WIMI)

[StockpickerCAPC]

WIMI Develops A 3D-Imaging Reconstruction Algorithm System.

Source
https://www.newstrail.com/wimi-develops-a-3d-imaging-reconstruction-algorithm-system/

February 9, 2023

With the continuous development of computer vision and spatial information technology, three-dimensional scene reconstruction is an important way to obtain spatial information. The fields involved in 3D scene reconstruction include computer science, data fusion, image processing, and other disciplines, and they cross each other. At present, 3D reconstruction technology is playing an increasingly important role in various fields, including geographical architecture, electronic mapping, film and television production, game development and innovation, and other fields. For instance, the producers can take the advantage of 3D reconstruction to shape a lot of fantasy and lifelike scenes, which make the audiences an immersive visual experience. Furthermore, with the popularity of the digital-city concept in recent years, the construction of a city also needs the support of 3D scene reconstruction technology while building the 3D smart city model.

3D reconstruction is a mathematical model that can establish suitable representation and processing for 3D objects. It is the base of the Process, Operation, and Analysis of 3D objects under the computer environment, and also, the key technology that expresses objective reality in virtual-reality.
It is understood that WIMI Hologram Cloud (NASDAQ: WIMI) has developed a 3D reconstruction algorithm system. It extracts the point cloud information from multiple images and the obtained point cloud information for data registration and fusion. In addition, the 3 d model reconstruction based on the fusion of point cloud information and image information can give full play to the advantages of the two kinds of information and provide accurate and rich realistic model information.

WIMI three-dimensional reconstruction algorithm system based on image and point cloud fusion mainly includes three parts:
-(1)Data acquisition program to obtain the multi-view cloud and image information of the target model;
-(2)Information extraction program, for sequence images can achieve feature extraction and matching, camera matrix calculation, dense reconstruction, and other functions;
- 3)Point cloud processing program, point cloud registration, and heterologous point cloud fusion.
The theoretical transition matrix is precalculated based on the load view information. The motion information extracted from the images was used to assist the point clouds for registration, followed by the fusion of the heterologous point clouds.

The purpose of point cloud registration is to align two point clouds of the same target acquired at different directions in the same environment, so as to obtain complete environment model information. In the registration process, the initial information is obtained by multi-point cloud and image acquisition of environmental targets.

After obtaining the initial information, the motion information is calculated from the image sequence. After dense reconstruction, the environmental 3D point cloud is obtained, and the calibration of the scale factor is completed in the process. Then the final environmental point cloud model is obtained after merging the point clouds, and the 3D scene model is obtained after surface reconstruction.
Among them, the information extraction stage, mainly involves feature extraction, matching, and acquisition of feature points and reference calibration outside the camera. On the one hand, the external parameter calibration of the camera mainly uses the extracted feature points to solve the information; on the other hand, the number and accuracy of feature point pairs are the basis of the subsequent reconstruction process. After collecting or merging to obtain the point cloud data, it is necessary to use appropriate algorithms to preprocess the collected point cloud data, and then surface-reconstruct the processed point cloud. The dense reconstruction is based on the dense matching of sparse point clouds, and further uses the image information to obtain rich point clouds.

3d scene reconstruction technology development has become an important computer technology, from a single object to multiple objects, from small targets to large target,s and even outdoor scenes, for 3 d target scene color information and space distance information acquisition and fusion has achieved very high accuracy, it’s in construction engineering, intelligent robot, entertainment games, and other fields have been widely used. WIMI will also continue to expand the application field of its three-dimensional reconstruction algorithm system based on image and point cloud fusion, build a bridge between real scenes and virtual space, and help shape the real digital world.