Near-far scene distributed parallel rendering method and system of aviation simulator

Abstract

The invention provides a near-far scene distributed parallel rendering method and system of an aviation simulator. The method comprises the following steps: dividing a whole rendering picture into a near-scene picture layer and a far-scene picture layer; independently rendering the distant view layer on a distant view GPU; dividing the distant view layer into odd frames and even frames, and rendering the odd frames and the even frames on different close-range GPUs; and performing frame-by-frame synthesis, segmentation and correction on the rendered long-range image and close-range image on a single image processing GPU, and restoring the corrected image into an image on a spherical surface according to a fixed frame frequency. A multi-channel hardware frame synchronization technology is abandoned, so that the hardware cost of the system is reduced, and a new thought is provided for research, development and innovation of an imaging system of the aviation simulator.

Description

technical field [0001] The invention relates to a visual system of an aviation simulator, in particular to a distributed parallel rendering method and system for far and near scenes of an aviation simulator. Background technique [0002] At present, mainstream visual IG usually generates and renders scenes according to sub-channels to realize real-time image generation of continuous large scenes. Its structure is as follows: figure 1 As shown, the specific principle is: after the scene management computer receives the position and attitude information of the flight simulation computer, it divides the entire rendering picture into left-eye, right-eye and plane rendering areas corresponding to different viewing angles according to the position and attitude information, and then The rendering instructions of different viewing angles are sent to each rendering computer through the network, and each rendering computer stores a set of identical visual databases. Each rendering co...

AI Technical Summary

Problems solved by technology

[0003] However, this method needs to store the same scene database locally on each rendering computer, which limits the complexity of the scene that the entire system can carry. Once the complexity of the scene exceeds the limit, the entire system will be difficult to run smoothly

In addition, the hardware frame synchronization technology used in this method has high requirements for hardware synchronization processing, and the hardware synchronization equipment that can meet this requirement is very expensive, making the cost of the entire visual IG system high

Images