CN114926604A - Collision detection method and system in three-dimensional scene - Google Patents

Abstract

The invention discloses a method and system for collision detection in a three-dimensional scene. The method for collision detection in a three-dimensional scene includes: acquiring an accessible area in the three-dimensional scene, and setting internal parameters of the system; The collision determination between the model position and the camera position is performed; if the player model is about to collide, the target position of the player model is recalculated; if the camera is about to collide, the target position of the camera is recalculated to realize the processing after the collision between the player model and the camera. .

Description

A method and system for collision detection in a three-dimensional scene

technical field

The present invention relates to the technical field of three-dimensional scene application, in particular to a method and system for collision detection in a three-dimensional scene.

Background technique

In 3D scene applications, in order to improve the simulation effect, a physical system is usually added to the application; the physical system can provide effective collision detection, calculation of gravity, additional force action, rigid body structure and other physical functions. However, the physical module itself is developed in a large volume, and a large number of verification algorithms will consume the performance of the host, so the cost of introducing the physical system is considerable. At the same time, the physical system requires the model to carry physical data, which will also increase the workload of data producers. However, many applications only need simple collision detection, so it is not suitable to introduce the whole set of physics system.

SUMMARY OF THE INVENTION

In view of this, the present invention proposes a method and system for collision detection in a three-dimensional scene. Based on the physical system, the collision judgment of the player model and the camera is realized, and the relevant processing after the collision between the player model and the camera is completed, and the lightweight system is realized. Collision detection function based on physics system.

A first aspect of the present invention provides a collision detection method in a three-dimensional scene. The collision detection method in a three-dimensional scene includes: acquiring an accessible area in a three-dimensional scene, and setting internal parameters of a physical system, including player position, player moving speed, and camera distance ; Calculate the target position of the player model and the camera separately, traverse the accessible area, and perform collision determination on the target position of the player model and the camera.

Further, the step of acquiring the accessible area in the three-dimensional scene includes: acquiring the main model data of the road network pavement in the three-dimensional scene; traversing all the triangular surfaces in the main model of the road network pavement, and collecting all the triangular surfaces facing upwards into one. A list of triangles to use as an accessible area.

Further, the step of traversing all the triangular surfaces in the main model of the road network surface includes: using the coordinates of the three vertices of the currently traversed triangular surface to determine the vectors of the two sides of the triangular surface; The vectors of the sides are cross-multiplied in order to obtain the normal vector N of the triangular surface; set the upward vector to U, and find the dot product M of the vector U and the normal vector N of the triangular surface; judge whether the dot product M is greater than 0, if the point If the product M is greater than 0, then the normal vector N and the vector U of the triangular face are facing the same, and the triangular face is judged to be facing upward.

Further, the step of calculating the target position of the player model includes: obtaining the current position of the player model and the current orientation of the player model; Time difference, calculate the target position expected by the player model.

Further, the step of traversing the accessible area to determine the collision of the target position of the player model: forming the first ray according to the expected target position of the player model and the vertical downward direction vector in the three-dimensional scene; traversing the triangle of the accessible area. list, and perform an intersection operation with the first ray to determine whether any triangles in the triangle list of the accessible area intersect the first ray; if any triangles in the triangle list of the accessible area intersect the first ray, it indicates that the player model expects The target position is in the accessible area and no collision has occurred; otherwise the player model is about to collide.

Further, the collision detection method in the three-dimensional scene further includes: if the player model is about to collide, recalculating the target position of the player model; the step of recalculating the target position of the player model includes: converting the expected target position of the player model to the target position. The direction of the player model is gradually stepped, and then a new target position is obtained. The stepping unit is a set player model body position; each time a step is made, the new target position of the player model is checked for collision. If a collision occurs, continue to step. Advance until the player model does not collide, and set the target position corresponding to the player model without collision as the final position of the player model.

Further, the step of calculating the target position of the camera includes: obtaining the calculated target position of the player model; obtaining the target orientation of the camera; according to the expected target position of the player model, the target orientation of the camera and the set camera distance, Calculate the target position expected by the camera.

Further, the step of traversing the accessible area to determine the collision of the target position of the camera: forming a second ray according to the target position expected by the camera and the vertical downward direction vector in the three-dimensional scene; traversing the triangle list of the accessible area, And perform an intersection operation with the second ray to determine whether there is a triangle in the triangle list of the accessible area that intersects the second ray; if there is a triangle in the triangle list of the accessible area that intersects with the second ray, it indicates the expected target position of the camera In the accessible area, there is no collision; otherwise the camera is about to collide.

Further, the collision detection method in the three-dimensional scene further includes: obtaining the final position of the player model, the expected target position of the camera, and the target orientation of the camera; stepping the expected target position of the camera along the target orientation of the camera, and the stepping unit is one. The set player model body position; the target position of the camera is checked for collision every step, if a collision occurs, the step will continue, otherwise the new target position of the camera will be obtained; it is judged whether the new target position of the camera is at the final position of the player model and the camera's expected target position; if the camera's new target position is between the player model's final position and the camera's expected target position, the camera's new target position is used as the camera's final position; otherwise, the camera's final position is set to the player's final position. The final position of the model.

A second aspect of the present invention provides a collision detection system in a three-dimensional scene. The collision detection setting in the three-dimensional scene includes: a memory for storing a computer program; and a processor for implementing the above-mentioned three-dimensional scene when the computer program is executed. The steps of the collision detection method.

The collision detection method in the above-mentioned three-dimensional scene uses an intersection algorithm between the target position point of the player model and the camera and the data of the accessible area to determine the occurrence of the collision, and completes the relevant processing after the collision between the player model and the camera, and recalculates The final position of the player model and the camera, to implement the collision detection function based on the lightweight physics system.

Description of drawings

For purposes of illustration and not limitation, the present invention will now be described in accordance with preferred embodiments thereof, particularly with reference to the accompanying drawings, wherein:

1 is a flowchart of a method for collision detection in a three-dimensional scene provided by an embodiment of the present invention;

FIG. 2 is a schematic diagram of a collision detection system in a three-dimensional scene provided by an embodiment of the present invention.

Detailed ways

In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments may be combined with each other under the condition of no conflict.

In the following description, many specific details are set forth in order to facilitate a full understanding of the present invention, and the described embodiments are only some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terms used herein in the description of the present invention are for the purpose of describing specific embodiments only, and are not intended to limit the present invention.

FIG. 1 is a flowchart of a method for collision detection in a three-dimensional scene provided by an embodiment of the present invention. The collision detection method in the three-dimensional scene judges the occurrence of a collision by performing an intersection algorithm between the target position point of the player model and the camera and the accessible area data, and completes the relevant processing after the collision between the player model and the camera, and recalculates the player model. and the final position of the camera.

Referring to Figure 1, the collision detection method in the 3D scene includes the following steps:

S100, acquiring an accessible area in the three-dimensional scene, and setting internal parameters of the physical system.

In this embodiment, the specific implementation of acquiring the accessible area in the three-dimensional scene and setting the internal parameters of the system is as follows:

S101, obtain an accessible area in a three-dimensional scene.

Obtain the main model data of the road network pavement in the 3D scene, and determine all the triangles facing upward in the main model of the road network pavement; then traverse all the triangular surfaces in the main model of the road network pavement, and collect them as a triangle list L, as an accessible area.

Among them, the specific implementation method of judging the upward facing of the triangular surface in the main model of the road network pavement is as follows:

Assuming that the three vertices composed of the currently traversed triangular faces are P0(x0,y0,z0), P1(x1,y1,z1), P2(x2,y2,z2) in counterclockwise order, we can get two edges The vector of is:

L01=P1-P0=(x1-x0, y1-y0, z1-z0)

L02＝P2-P0＝(x2-x0,y2-y0,z2-z0)

Cross-multiply the two edges in order to get the normal vector Normal (Nx, Ny, Nz) of the triangular face. The normal vector Normal(Nx, Ny, Nz) of the triangular face is:

Normal=cross(L01,L02)=(

(y1-y0)(z2-z0)-(z1-z0)(y2-y0),

(z1-z0)(x2-x0)-(z2-z0)(x1-x0),

(x1-x0)(y2-y0)-(y1-y0)(x2-x0)

)

Set the vertical upward direction vector in the scene to U(0,1,0), and find the dot product Mark of the vector U and the normal vector Normal of the triangular surface. The dot product Mark of U and Normal is:

Mark=Dot(U,Normal)=Ux*Nx+Uy*Ny+Uz*Nz=Ny=(z1-z0)(x2-x0)-(z2-z0)(x1-x0)

From the mathematical meaning of the dot product, when the mark is greater than 0, the Normal and U are in the same direction, and it can be judged that the triangular face is facing upward.

Among them, the mathematical meaning of dot product: Dot(A, B)=|A||B|cosα, α represents the angle between A and B.

In an embodiment, in order to optimize performance, only (z1-z0)(x2-x0)-(z2-z0)(x1-x0) is calculated to determine whether the triangular faces are facing upward.

S102, the internal parameters of the physical system are set.

In this embodiment, the set internal parameters of the system include the player's position, the player's moving speed, and the camera distance.

Among them, the player's body (body_width): Set the width of the player's body to 0.4 meters. This value will be used for the step unit of the player model's position adjustment when a collision occurs.

Among them, the player's movement speed (body_speed): Set the player's movement speed to 5 meters per second, which is used to control the speed of the player's model movement.

Among them, the camera distance (camera_dis): set the camera distance to 3 meters, this value represents the distance between the camera position and the player model when there is no collision.

S200: Perform collision determination on the given position P(x, y, z) and the accessible area obtained in step S100.

In this embodiment, the specific implementation manner of performing collision determination on the given position P(x, y, z) and the accessible area obtained in step S100 is as follows:

The collision judgment is performed on the given position P(x, y, z) and the accessible area calculated in step S100, and the judgment method is as follows:

S201, according to the given position P(x, y, z) and the vertical downward direction vector Dir(0, -1, 0) in the current three-dimensional scene, form a ray Ray(P, dir).

S202, traverse the triangle list L of the accessible area, perform an intersection operation with the ray Ray, and determine whether there is a triangle (p0, p1, p2) in the triangle list L of the accessible area that intersects the ray Ray.

In this embodiment, the specific implementation method for judging that there is a triangle (p0, p1, p2) intersecting with the ray Ray in the triangle list L of the accessible area is as follows:

(1) Assuming a point is on both the ray (P, dir) and the triangle (p0, p1, p2), the following expression is satisfied:

P+t*dir=(1-u-v)*p0+u*p1+v*p2

Converted to:

(p1-p0)*u+(p2-p0)*v-t*dir=P-p0

Among them, t, u, v are coefficients respectively.

(2) In order to rewrite the above converted expression into matrix form, set:

E0=p1-p0

E2=p2-p0

E3=-dir

A=(E0,E1,E2)

B=P-p0

Then there are:

A*X=B

which is:

(3) According to Kram's rule, the expressions of coefficients u, v and t are obtained as:

u=det(B,E1,E2)/det(A)

v=det(E0,B,E2)/det(A)

t=det(E0,E1,B)/det(A)

(4) Optimization coefficient

According to the characteristic that the mixed product represents the same volume as the determinant under the three-dimensional vector, the parameters can be optimized to obtain:

det(A)=dot(E0,cross(E1,E2))

u=dot(cross(B,E1),E2)/det(A)

v=dot(cross(E0,B),E2)/det(A)

t=dot(cross(E0,E1),B)/det(A)

(5) According to the optimization parameters, determine whether the ray Ray (P, dir) intersects the triangle (p0, p1, p2).

If u>=0, v>=0, u+v<=1 and t>=0, it is determined that the ray Ray(P, dir) intersects the triangle (p0, p1, p2), otherwise it does not intersect.

The principle of this embodiment for judging whether the ray (P, dir) intersects the triangle (p0, p1, p2) is as follows:

(1) If a point is in the triangle (p0, p1, p2), it can be represented by (1-u-v)*p0+u*P1+v*p2, at this time u>=0, v>=0, u+v<=1.

(2) If a point is on the ray (P, dir), it can be represented by P+t*dir, t>=0.

S203 , if a triangle (p0, p1, p2) in the triangle list L of the accessible area intersects with the ray Ray, it means that the given position P is in the accessible area and no collision occurs; otherwise, a collision occurs.

S300, based on the aforementioned collision determination method, sequentially perform collision determination on the player position and the camera position in the three-dimensional scene.

In this example, based on the above-mentioned collision determination method, the collision determination of the player position and the camera position in the three-dimensional scene includes the following steps:

S301, player position collision determination.

When the player model moves, the specific implementation method of the collision determination on the position of the player model is as follows:

S301-1, obtain the current position pos of the player model and the current orientation dir of the player model.

S301-2, according to the current position pos of the player model, the current direction of the player model dir, the set player movement speed body_speed, and the time difference delta between the current frame and the previous frame, calculate and obtain the target position target predicted by the player model.

In this embodiment, the calculation method of the target position target estimated by the player model is as follows:

Target=pos+(delta*body_speed)*dir

S301-3, according to the collision method in step S200, it is determined whether the player model is about to collide according to the target position target predicted by the player model.

In this embodiment, the step of determining whether the player model is about to collide by using the target position target predicted by the player model includes:

(1) The first ray Ray1(T, dir) is formed according to the target position target predicted by the player model and the vertical downward direction vector Dir(0,-1,0) in the current three-dimensional scene.

(2) Traverse the triangle list L of the accessible area, and perform an intersection operation with the first ray Ray1 to determine whether there is a triangle (p0, p1, p2) in the triangle list L of the accessible area and the first ray Ray1 (T, dir) to intersect.

In this embodiment, the step of judging whether a triangle (p0, p1, p2) intersects with the first ray Ray1 (T, dir) in the triangle list L of the accessible area includes:

(2-1) Assuming that a point is on both the ray (T, dir1) and the triangle (p0, p1, p2), the following expression is satisfied:

T+t1*dir1=(1-u1-v1)*p0+u1*p1+v1*p2

Converted to:

(p1-p0)*u1+(p2-p0)*v1-t1*dir=T-p0

Among them, t1, u1, v1 are coefficients respectively.

(2-2) In order to rewrite the above converted expression into matrix form, set:

E0=p1-p0

E2=p2-p0

E3=-dir1

A1=(E0,E1,E2)

B1=T-p0

Then there are:

A1*X1=B1

which is:

(2-3) According to Kram's rule, the expressions for the coefficients u1, v1, and t1 are:

u1=det(B1,E1,E2)/det(A1)

v1=det(E0,B,E2)/det(A1)

t1=det(E0,E1,B1)/det(A1)

(2-4) Optimization coefficient

According to the characteristic that the mixed product represents the same volume as the determinant under the three-dimensional vector, the parameters can be optimized to obtain:

det(A1)=dot(E0,cross(E1,E2))

u1=dot(cross(B1,E1),E2)/det(A1)

v1=dot(cross(E0,B1),E2)/det(A1)

t1=dot(cross(E0,E1),B1)/det(A1)

(2-5) According to the optimization parameters, determine whether the first ray Ray1 (T, dir1) intersects the triangle (p0, p1, p2).

If u1>=0, v1>=0, u1+v1<=1 and t1>=0, it is determined that the first ray Ray(T, dir1) intersects with the triangle (p0, p1, p2), otherwise they do not intersect.

(3) If there is a triangle (p0, p1, p2) in the triangle list L of the accessible area that intersects the first ray Ray1 (T, dir1), it means that the target position target predicted by the player model is in the accessible area, and no collision occurs ; otherwise a collision occurs.

In this embodiment, the collision detection of the player model is realized by performing an intersection algorithm between the target position point of the player model and the data of the accessible area to determine the occurrence of the collision of the player model.

S302, camera position collision determination

In this embodiment, the referred camera is a camera outside the player. The collision judgment result of the camera of the player model and the collision judgment result of the player's position do not need to be processed separately.

When the player moves or rotates, the specific implementation of the collision determination of the camera position is as follows:

S302-1, acquiring the target position target predicted by the player model calculated in step S301.

S302-2, obtain the target orientation dir of the camera.

S302-3, according to the target position target of the player, the target orientation dir of the camera, and the set camera distance camera_dis, calculate the target position camera_target expected by the camera.

In this embodiment, the calculation method of the target position camera_target estimated by the camera is as follows:

Camera_target=target-camera_dis*dir

S302-4, according to the collision method in step S200, it is determined whether the camera is about to collide according to the target position camera_target expected by the camera.

In this embodiment, the specific implementation manner of determining whether the camera is about to collide by using the camera_target estimated target position of the camera is as follows:

(1) A second ray Ray2(C, dir2) is formed according to the target position camera_target predicted by the camera and the vertical downward direction vector Dir(0,-1,0) in the current three-dimensional scene.

(2) Traverse the triangle list L of the accessible area, and perform an intersection operation with the second ray Ray2 to determine whether there are triangles (p0, p1, p2) and the second ray (C, dir2) in the triangle list L of the accessible area. intersect.

In this embodiment, the step of judging whether a triangle (p0, p1, p2) intersects with the second ray Ray2 (C, dir2) in the triangle list L of the accessible area includes:

(2-1) Assuming that a point is on both the second ray Ray2 (C, dir2) and the triangle (p0, p1, p2), the following expression is satisfied:

C+t2*dir2=(1-u2-v2)*p0+u2*p1+v2*p2

Converted to:

(p1-p0)*u2+(p2-p0)*v2-t2*dir2=C-p0

Among them, t2, u2, v2 are coefficients respectively.

(2-2) In order to rewrite the above converted expression into matrix form, set:

E0=p1-p0

E2=p2-p0

E3=-dir2

A2=(E0,E1,E2)

B2=C-p0

Then there are:

A2*X2=B2

which is:

(2-3) According to Kram's rule, the expressions of coefficients u2, v2 and t2 are obtained as:

u2=det(B2,E1,E2)/det(A2)

v2=det(E0,B2,E2)/det(A2)

t2=det(E0,E1,B2)/det(A2)

(2-4) Optimization coefficient

According to the characteristic that the mixed product represents the same volume as the determinant under the three-dimensional vector, the parameters can be optimized to obtain:

det(A2)=dot(E0,cross(E1,E2))

u2=dot(cross(B1,E1),E2)/det(A2)

v2=dot(cross(E0,B2),E2)/det(A2)

t2=dot(cross(E0,E1),B2)/det(A2)

(2-5) According to the optimization parameters, determine whether the second ray Ray2 (C, dir2) intersects the triangle (p0, p1, p2).

If u2>=0, v2>=0, u2+v2<=1 and t2>=0, it is determined that the second ray Ra2y(C, dir2) intersects the triangle (p0, p1, p2), otherwise it does not intersect.

(3) If there is a triangle (p0, p1, p2) in the triangle list L of the accessible area and the second ray Ray2 (C, dir2) intersects, it means that the camera's expected target position camera_target is in the accessible area, and there is no collision; Otherwise a collision occurs.

In this embodiment, the camera collision detection is realized by performing an intersection algorithm between the target position point of the camera and the data of the accessible area to determine the occurrence of the camera collision.

S400, if the player model is about to collide, recalculate the target position of the player model.

In this embodiment, the step of recalculating the target position of the player model includes:

Stepping the expected target position target of the player model obtained in step S301 toward the player model gradually, and then obtaining a new target position temp, and the stepping unit is a set player model position; If there is a collision, continue stepping, otherwise temp is the final target position of the player model. Finally, set the player model position to temp.

In this embodiment, when the player model is about to collide, the collision processing is performed on the player model, and the target position of the player model is recalculated, so as to realize the collision detection of the player model based on the basis of the physics system.

S500, if the camera is about to collide, recalculate the target position of the camera.

In this embodiment, the camera referred to is the camera outside the player model. There is no need for additional collision processing for the camera of the player model, just wait for the player collision to be processed, and directly assign the camera to the player's position and the player's orientation.

In this embodiment, the collision processing of the camera is placed after the collision processing of the player.

In this embodiment, if the camera is about to collide, the specific implementation method of recalculating the target position of the camera is as follows:

Obtain the final position pos of the player, the expected target position camera_target of the camera obtained in step S302, and the target of the camera is oriented toward dir;

Step the camera's expected target position camera_target along the camera's target towards dir, the step unit is a set player model position; each step is a collision judgment, if a collision occurs, continue to step, otherwise get The camera's new target position temp.

Determine whether the new target position temp of the camera is located between the final position pos of the player model and the expected target position camera_target of the camera obtained in step S302;

If the new target position temp of the camera is located between the final position pos of the player model and the expected target position camera_target of the camera obtained in step S302, the final position of the camera is set to temp, otherwise, the final position of the camera is set to the final position of the player model pos.

In this embodiment, the method for judging whether the new target position temp of the camera is located between the final position pos of the player model and the expected target position camera_target of the camera obtained in step S302 is as follows:

Set the camera's new target position temp to P, the player model's final position pos to A, and the camera's expected target position camera_target to B.

Two position vectors are obtained, namely PA and PB, where PA=A-P; PB=B-P.

Calculate the cosine value Cosα of the two position vectors, where the calculation method of Cosα is:

Cosα=dot(PA, PB)/(|PA|*|PB|)

If the cosine value Cosα of the two position vectors is equal to 1, it means that P is between AB, that is, the new target position temp of the camera is located between the final position pos of the player model and the expected target position camera_target of the camera. Otherwise, P is not between AB.

In this embodiment, when the camera is about to collide, the camera is subjected to collision processing, the target position of the camera is recalculated, and the camera collision detection based on the basis of the physical system is realized.

The collision detection method in the above-mentioned three-dimensional scene uses an intersection algorithm between the target position point of the player model and the camera and the data of the accessible area to determine the occurrence of the collision, and completes the relevant processing after the collision between the player model and the camera, and recalculates The final position of the player model and the camera, to implement the collision detection function based on the lightweight physics system.

The above-mentioned collision detection method in a three-dimensional scene can provide complete collision detection and related functions of player model and camera collision processing without adding a large amount of physical data.

The above-mentioned collision detection method in a 3D scene is very lightweight, and because the main model data of the road network is usually relatively simple, it will not bring pressure on performance, and it is convenient to introduce into various applications.

Corresponding to the above method embodiments, see FIG. 2 , which is a schematic diagram of a collision detection system in a three-dimensional scene provided by another embodiment of the present invention. The device 100 may include:

memory 101 for storing computer programs;

The processor 102 can implement the following steps when executing the computer program stored in the above-mentioned memory 101:

Obtain the accessible area in the 3D scene, and set the internal parameters of the system; perform a collision determination between the given position P(x, y, z) and the acquired accessible area; based on the above collision determination method, sequentially determine the player's position in the 3D scene Determine the collision with the camera position; if the player model is about to collide, the target position of the player model is recalculated; if the camera is about to collide, the target position of the camera is recalculated to realize the processing after the collision between the player model and the camera.

The above-mentioned specific embodiments do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims (10)

1. a collision detection method in a three-dimensional scene, is characterized in that, comprises: Get the accessible area in the 3D scene, and set the internal parameters of the physical system, including the player's position, the player's movement speed, and the camera distance; Calculate the target positions of the player model and the camera respectively, traverse the accessible area, and perform collision determination on the target positions of the player model and the camera. 2. The collision detection method in a three-dimensional scene according to claim 1, wherein the step of acquiring the accessible area in the three-dimensional scene comprises: Obtain the main model data of the road network pavement in the 3D scene; Traverse all the triangular faces in the main model of the road network, collect all the triangular faces facing upward as a triangular list, and use the triangular list as an accessible area. 3. The collision detection method in a three-dimensional scene according to claim 2, wherein the step of traversing all the triangular surfaces in the main model of the road network surface comprises: Use the coordinates of the three vertices of the currently traversed triangular face to determine the vectors of the two sides of the triangular face; Cross-multiply the vectors of the two sides of the triangular face in order to obtain the normal vector N of the triangular face; Set the upward vector as U, and find the dot product M of the vector U and the normal vector N of the triangular surface; Determine whether the dot product M is greater than 0. If the dot product M is greater than 0, the normal vector N and the vector U of the triangular face are facing the same, and the triangular face is judged to face upward. 4. The collision detection method in a three-dimensional scene according to claim 1, wherein the step of calculating the target position of the player model comprises: Get the current position of the player model and the current orientation of the player model; Calculate the expected target position of the player model according to the current position of the player model, the current orientation of the player model, the set player movement speed, and the time difference between the current frame and the previous frame. 5. The method for detecting collision in a three-dimensional scene according to claim 4, wherein the step of traversing the accessible area and judging the target position of the player model for collision: The first ray is formed according to the target position predicted by the player model and the direction vector in the current 3D scene; Traverse the triangle list of the accessible area, and perform an intersection operation with the first ray to determine whether any triangles in the triangle list of the accessible area intersect the first ray; If a triangle in the triangle list of the accessible area intersects with the first ray, it means that the expected target position of the player model is in the accessible area, and no collision has occurred; otherwise, the player model is about to collide. 6. The collision detection method in a three-dimensional scene according to claim 5, further comprising: If the player model is about to collide, recalculate the target position of the player model; the step of recalculating the target position of the player model includes: Gradually step the expected target position of the player model towards the direction of the player model, and then obtain a new target position. The stepping unit is a set player model body position; each step, the new target position of the player model is checked for collision. , if a collision occurs, continue stepping until the player model does not collide, and set the target position corresponding to the player model without collision as the final position of the player model. 7. The collision detection method in a three-dimensional scene according to claim 4, wherein the step of calculating the target position of the camera comprises: Get the calculated target position of the player model; Get the target orientation of the camera; Calculate the expected target position of the camera according to the target position expected by the player model, the target orientation of the camera, and the set camera distance. 8. The method for detecting collision in a three-dimensional scene according to claim 7, wherein the step of traversing the accessible area and judging the target position of the camera for collision: A second ray is formed according to the target position predicted by the camera and the direction vector in the current 3D scene; Traverse the triangle list of the accessible area, and perform an intersection operation with the second ray to determine whether any triangles in the triangle list of the accessible area intersect with the second ray; If a triangle in the triangle list of the accessible area intersects with the second ray, it means that the target position expected by the camera is in the accessible area, and no collision occurs; otherwise, the camera is about to collide. 9. The method for detecting collision in a three-dimensional scene according to claim 8, further comprising: Obtain the final position of the player model, the expected target position of the camera, and the target orientation of the camera; Step the camera's expected target position along the camera's target direction, and the step unit is a set player model's body position; the target position of the camera is checked for collision at each step. If a collision occurs, continue to step. Otherwise, get the new target position of the camera; Determine whether the new target position of the camera is between the final position of the player model and the expected target position of the camera; If the new target position of the camera is between the final position of the player model and the expected target position of the camera, the new target position of the camera is used as the final position of the camera; otherwise, the final position of the camera is set as the final position of the player model. 10. A collision detection system in a three-dimensional scene, comprising: memory for storing computer programs; The processor is configured to implement the steps of the collision detection method in a three-dimensional scene according to any one of claims 1 to 9 when executing the computer program.

Images