Experts talk about how to use 3D point cloud technology to make the VR experience more realistic

Tice is the Executive Director of 3D Visualization and Design Company ToPa 3D. A few days ago, he published an article on l inkedIn related to point cloud data used by Australian technology company Euclideon, and caused some discussion in the industry.

Point cloud problem

Point clouds are data collected by 3D scanning hardware such as FARO's Focus 3D laser scanner and Shining 3D's Einscan Pro. The basic principle of capturing a 3D object is a single point at which the 3D scanner feeds back the beam to the surface. Even if you scan the smallest objects, the system will create millions of such points, so managing such massive data is not easy. CAD software can connect to various points, but this process requires a lot of computing resources, and perfecting the finished model is often a difficult process.

From the point cloud (right) to the completed 3D model (left and center)

The unique infinite detail algorithm is integrated into the SOLIDSCAN software to render the point cloud in real time for interactive 3D visualization of the scan. The company said, "This algorithm only captures one point in each screen pixel, not every small point."

When citing Eugene's technology, Tice's following statement has caused discussion in the industry:

If Youli can animate the point cloud as the video demonstrates, then it should be possible to assign metadata to each point, which can be hosted as an IP in the Internet of Things. If this is possible, ToPa believes that one day, component and mesh modeling will be eliminated because future smart models will consist of pure point cloud data. ”

Here Tice predicts two potential directions for the "point cloud future": collecting metadata along with the point cloud; an intelligent 3D model independent of the grid.

2. In the Internet of Things

To understand the first meaning, feeding metadata to a smart IoT system can help the machine react better to certain conditions. Information such as location, color, texture, scale, and how a point relates to other points can help advances in areas such as augmented reality, virtual reality, and the like.

When HTC Vive went on the market, VR became a hot topic for many technology companies. To prove that VR exists in real-world applications other than games, 3D software giant Dassault Systèmes integrates HTC Vive into its 3DEXPERIENCE product viewer, enabling users to immerse 3D designs.

As for the "intelligent 3D model independent of the grid", this will enable faster and more reliable 3D visualization access. Youli cited the 3D Google Earth type of image as a future development for this:

Imagine a 3D landscape that is large enough to extend to the horizon, but you can zoom in on a particle scattered over the ground in less than a second.

We don't know what you think, but we absolutely hope that we can walk on the moon. The views drawn by Tice also attracted people's attention:

Imagine using LiDAR to investigate the entire planet and observe it in real time.

This is definitely an exciting prospect for the players. Combining sparse point cloud data into surfaces instead of polygons provides a more realistic virtual world and reduces the load on the host's computation.

Ningbo Autrends International Trade Co.,Ltd. , https://www.vapee-cigarettes.com