CN-121982179-A - Real-time rendering optimization method for three-dimensional animation special effects

CN121982179ACN 121982179 ACN121982179 ACN 121982179ACN-121982179-A

Abstract

The invention provides a real-time rendering optimization method of a three-dimensional animation special effect, which comprises the steps of dividing the special effect into three types, namely a static attached special effect, a dynamic tracking special effect and an environment interaction special effect, determining the rendering priority and detail level of the static attached special effect, determining whether to perform view cone rejection or delay rendering on the dynamic tracking special effect, determining a detail level model adopted by the environment interaction special effect, performing predictive asynchronous loading and unloading on special effect resources in response to changes of game progress or scene states, monitoring rendering performance in real time, and dynamically adjusting the optimizing strategy strength aiming at different special effect types according to monitored performance indexes. The invention can maximize the balance between the rendering efficiency and the visual fidelity under the condition of limited hardware resources, and enhance the applicability of the scheme on different performance platforms.

Inventors

HU JIANKAI
Zhu Xiongxi
LIU XI
HONG FUCHENG
CHENG TENG
ZHENG ZONGLIN
ZHANG ZHIWEI

Assignees

深圳市代宝科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260224

Claims (10)

1. A method for optimizing real-time rendering of three-dimensional animated special effects, which is applied to an electronic game, and is characterized by comprising the following steps: According to the dynamic interaction relation between the special effect resource and the target object in the game scene, dividing the special effect into three types, namely a static attached special effect, a dynamic tracking special effect and an environment interaction special effect; Determining the rendering priority and the detail level of the static attached special effect according to the visual importance of the target object attached to the static attached special effect in a game picture; determining whether to perform view cone rejection or delay rendering on the dynamic tracking special effect according to the relative positions and the motion states of the dynamic tracking special effect and the virtual camera; Determining a detail level model adopted by the environment interaction special effect according to the distance between the environment interaction special effect and the virtual camera; Responding to the change of game progress or scene state, carrying out predictive asynchronous loading and unloading on the special effect resource; And monitoring rendering performance in real time, and dynamically adjusting the optimized strategy strength aiming at different special effect types according to the monitored performance index.
2. The method according to claim 1, wherein said determining the rendering priority and level of detail of the static attached effect comprises: when the target object is a role controlled by a player or a main interaction object of the role, maintaining high-level-of-detail rendering of the static attached special effect; and when the target object is a scene background element or a non-key object, reducing or simplifying the detail level of the static attached special effect.
3. The method according to claim 1, wherein said determining whether to view-cone cull or delay render the dynamically tracked special effects comprises: for the dynamic tracking special effect of which the motion trail is about to enter a virtual camera view cone, preloading high-detail level resources of the dynamic tracking special effect; And rendering the dynamic tracking special effect which moves rapidly and has short screen residence time by adopting a motion blur simplification model.
4. The method according to claim 1, wherein said determining a level of detail model employed by said ambient interaction effect comprises: Enabling physical interaction simulation and high-particle number rendering for the environmental interaction special effects within a first short-distance threshold from the virtual camera; rendering with a simplified particle system or animation map for the environmental interaction special effects whose distance exceeds the first close-range threshold but does not exceed the second far-range threshold; and rendering the environmental interaction special effects with the distance exceeding the second long-distance threshold by adopting a static bulletin board or simplifying the static bulletin board into color blocks.
5. The method according to claim 1, wherein said predictive asynchronous loading and unloading of said effect resources comprises: according to predefined data of a game level map or player movement trend analysis, loading an environment interaction special effect resource package which possibly appears in a next game area into a memory in advance; Unnecessary special effects resource packages of which the player character is far away from the area are marked as an uninstallable state.
6. The method of claim 1, wherein the real-time monitoring of rendering performance and dynamically adjusting the optimization strategy strength for different effect types based on the monitored performance metrics comprises: Predefining a plurality of policy levels corresponding to the performance indicators, each policy level associated with a set of differentiated level of detail configurations for the static attachment effect, the dynamic tracking effect, and the environmental interaction effect; Comparing the monitored performance index with the threshold value of each strategy level in real time to determine the current strategy level; and dynamically adjusting the rendering detail levels of various special effects according to the configuration associated with the current policy level, wherein the detail levels of the static attached special effects and the dynamic tracking special effects are recovered preferentially when switching from a low performance level to a high performance level, and then the detail levels of the environment interaction special effects are recovered.
7. The method of claim 1, wherein the static attachment effect comprises a persistent state indication effect attached to a surface of the interactable object and a persistent defensive visual effect surrounding the target object; the dynamic tracking special effects comprise a projector special effect with an autonomous motion track and a track residual special effect generated by the rear of a moving object; The ambient interactive special effects include persistent ambient atmosphere special effects associated with scene fixation, dynamically diffused regional visual masking special effects, and interactive triggered surface disturbance special effects.
8. The method of claim 7, wherein the persistence state indicates a special effect, and wherein a rendering level of detail is dynamically adjusted based on a current interaction state of the interactable object; when the interactable object is in a high interactivity state, rendering a high level of detail is adopted for the persistence state indication special effect; And when the interactable object is in a low interactivity or idle state, adopting low-level-of-detail rendering for the persistence state indication special effect.
9. The method of claim 1, wherein the method is applied to a mobile terminal device, and wherein dynamically adjusting the optimized policy strength for different effect types further comprises: When the heating of the mobile terminal equipment or the reduction of the frame rate is detected, the global particle number and the physical simulation precision of the environment interaction special effect are preferentially reduced, and the simultaneous display number of the dynamic tracking special effect is limited.
10. The method of claim 1, wherein the method is applied to a virtual reality system having a head mounted display device, and wherein the determining whether to view-cone cull or delay render the dynamic tracking effect comprises: Based on the real-time gesture data of the head-mounted display device, predicting the view cone range of the virtual camera in the next frame, and enabling progressive detail loading for the dynamic tracking special effect in the predicted view cone edge area.

Description

Real-time rendering optimization method for three-dimensional animation special effects Technical Field The application relates to the technical field of computer graphics processing, in particular to a real-time rendering optimization method for a three-dimensional animation special effect. Background In electronic games, especially large three-dimensional games, real-time rendering is a core and computationally intensive task. With the increasing complexity of game scenes and the increasing requirements of players on visual quality, three-dimensional animation special effects (such as explosion, magic, flame, airflow and the like) play a key role in creating immersion, but the performance pressure brought by the special effects is increasingly remarkable. At present, various optimization techniques are commonly adopted in the industry to ensure the smoothness of real-time rendering, and mainly comprise a level of detail (LOD) model, a view cone rejection, an object rendering outside the view of a camera, asynchronous resource loading and the prevention of clamping caused by disk I/O, wherein the model with different precision is switched according to the distance between the object and the camera. However, these techniques have significant drawbacks when applied to massive and behavioural special effects. First, existing methods generally treat special effects as common visual objects, make optimization decisions using uniform criteria (e.g., distance), and ignore distinct behavior patterns of different special effects in game logic (e.g., attach to characters, track targets, interact with the environment). The optimization strategy of 'one-cut' leads to either excessive simplification of the visual key special effects due to the long distance, affecting the core game experience, or serious waste of graphics processing resources due to the reservation of a large number of unnecessarily high detail special effects. Secondly, the existing system lacks a dynamic classification and resource scheduling framework based on special effect behavior logic, and the strength and the priority of an optimization strategy cannot be intelligently and differentially adjusted according to scene state changes (such as combat explosion and scene switching) and real-time performance indexes (such as frame rate and GPU load) in the game running process. Finally, in performance or delay sensitive scenarios such as mobile devices, virtual Reality (VR), etc., it is often difficult for existing general schemes to maintain an optimal balance of special effect visual quality and fluency with limited computational effort and demanding delay requirements. Therefore, a more intelligent and more refined real-time rendering optimization method for three-dimensional animation special effects is needed in the art, which can understand the interactive intention of the special effects and implement self-adaptive and differentiated resource management and rendering strategies according to the interactive intention. Disclosure of Invention The invention aims to overcome the defects of the prior art, and provides a real-time rendering optimization method for a three-dimensional animation special effect, which aims to solve the technical problems that in the prior art, the special effect is optimized by 'one-cut' so as to cause unbalanced visual quality and rendering efficiency, intelligent classification and resource scheduling based on special effect behavior logic are lacked, and self-adaptive refined control is difficult to realize on different performance platforms. In a first aspect, the present invention provides a method for optimizing real-time rendering of a three-dimensional animated special effect, applied to an electronic game, the method comprising: According to the dynamic interaction relation between the special effect resource and the target object in the game scene, dividing the special effect into three types, namely a static attached special effect, a dynamic tracking special effect and an environment interaction special effect; Determining the rendering priority and the detail level of the static attached special effect according to the visual importance of the target object attached to the static attached special effect in a game picture; determining whether to perform view cone rejection or delay rendering on the dynamic tracking special effect according to the relative positions and the motion states of the dynamic tracking special effect and the virtual camera; Determining a detail level model adopted by the environment interaction special effect according to the distance between the environment interaction special effect and the virtual camera; Responding to the change of game progress or scene state, carrying out predictive asynchronous loading and unloading on the special effect resource; And monitoring rendering performance in real time, and dynamically adjusting the optimized strategy strength aiming at different sp