High-performance rendering method of three dimensional (3D) point sampling data

Abstract

The invention provides a high-performance rendering method of three dimensional (3D) point sampling data, comprising the following steps of: estimating normal direction information of a 3D sampling point by constructing a local implicit surface of the nearest neighborhood set of the 3D sampling point in combination with a differential geometry theory; determining oval expression of the 3D sampling point on a tangent plane through analyzing the neighborhood set via covariance; drawing the 3D sampling point data by oval weighting filtering with excellent anti-aliasing effect so as to acquire a data rendering effect with high quality; and designing a single-pass drawing algorithm of the 3D point sampling data according to order of depth values of the sampling points under the current visual angle so as to accelerate rendering speed. By using the method, the 3D sampling data are rendered in high performance and high quality.

Description

technical field [0001] The invention relates to the field of computer graphics processing, in particular to a high-performance rendering method for three-dimensional point sampling data. Background technique [0002] With the rapid development of 3D scanning acquisition technology, 3D point sampling data model has become a new digital media after one-dimensional sound data, two-dimensional image data and video data. In the fields of reverse engineering, innovative design of industrial products, digital entertainment, film and television animation, e-commerce, communication, physical simulation, protection and restoration of cultural relics, 3D point sampling data models have been widely used, and have produced more and more far-reaching influences . In recent years, a variety of modeling and modeling methods for 3D point sampling data have emerged, which makes high-performance, fast and high-quality rendering of 3D point sampling data a very important task. [0003] Since ...

AI Technical Summary

Problems solved by technology

[0004] Although the hardware-based EWA algorithm improves the drawing speed, there are corresponding shortcomings

First, it usually requires multiple passes of drawing. In the first pass of drawing, it is necessary to move the object by a small offset away from the viewing angle and then draw the depth buffer. However, the determination of the offset is not a matter of It's easy, too large or too small an offset will affect the drawing effect

Second, the multi-pass rendering algorithm only draws the points at the forefront of the viewing angle, so it is impossible to draw translucent objects realistically

Images