【转】3D引擎中的剪切算法

原创 2011年01月12日 11:21:00

摘自: ttp://www.cs.berkeley.edu/~ug/slide/pipeline/assignments/as6/discussion.shtml

University of California
EECS Dept, CS Division

TA: Laura Downs
TA: Jordan Smith
CS184: Foundations of
Computer Graphics

Prof. Carlo H. Séquin
Spring 1998


CS 184: Discussion Section 7 Notes

Transforms in the Hierarchy (an Example)

Suppose I have a scene hierarchy as shown at the right. The top level node, or world node, is a translated, scaled box. The bottom level node, or object node, is a box with sides of length and one corner at the origin.

The Qworld<-object transform is the transformation that computes the world coordinates of a point given in object coordinates, i.e. (in column vectors) Qworld<-objectpobject=pworld. In this example, if I take the point, pobject=(0 1 1), on one corner of the original box, and apply the two modeling transforms, the world coordinates of that point, pworld, should be (1 4 3).

In the figure, the world coordinate system is system A and the object coordinate system is system C. Therefore, our world<-object transform will be the same as the A<-C transform, i.e. Qworld<-object=QA<-C,   pobject=pC,  and pworld=pA.


QA<-B
T(1 2 1)
<-----

QB<-A
T(-1 -2 -1)
----->


QB<-C
S(2 2 2)
<-----

QC<-B
S(1/2 1/2 1/2)
----->

We can't get the QA<-C transform directly, but we can construct it from the individual transforms in the hierarchy. We know QA<-B and QB<-C;   QA<-B is the translation, T(1 2 1)   and QB<-C is the scale, S(2 2 2).   Now we have enough information to compute the complete world<-object transform,

Qworld<-objectpobject =QA<-CpC =QA<-BQB<-CpC =T(1 2 1)S(2 2 2)pC

Qworld<-object =T(1 2 1)S(2 2 2)=

1
0
0
1

0
1
0
2

0
0
1
1

0
0
0
1

2
0
0
0

0
2
0
0

0
0
2
0

0
0
0
1

=

2
0
0
1

0
2
0
2

0
0
2
1

0
0
0
1

We also need to know the inverse to that matrix, the Qobject<-world matrix. Rather than invert the Qworld<-object matrix explicitly, we can keep track of the inverse as we move down the hierarchy.

Qworld<-object-1 =Qobject<-world =QC<-A =QC<-BQB<-A = S(1/2 1/2 1/2) T(-1 -2 -1).

Camera Transform

The camera transformation computes the coordinates of a point in space relative to the camera. This coordinate system is called the View Reference Coordinates (VRC). This transformation is desirable because camera projections are very easy to compute in this coordinate system.

In the example shown on the right, we have a scene hierarchy containing an object and a camera. Each occurs deep within the hierarchy and is affected by a sequence of transformations. We want to find the camera<-object transformation.

We already know how to find the world<-object transformation that transforms the points in an object's coordinate system into points in the world coordinate system. This is just the concatenation of all of the transforms along the path from the world to the object.

pworld = Qworld<-objectpobject = O1O2...Ompobject
Qworld<-object = O1O2...Om

We also know how to compute the inverse of this matrix.

Qobject<-world = Qworld<-object-1 = [O1O2...Om]-1 = Om-1...O2-1O1-1

We keep track of both of these matrices as we traverse the scene hierarchy. Each time a transformation is encountered in the traversal, we concatenate it with the Qworld<-object matrix and we concatenate its inverse with the Qobject<-world matrix.

The camera<-object transform is computed from the camera<-world and world<-object transforms.

pcamera = Qcamera<-objectpobject = Qcamera<-worldQworld<-objectpobject

When we want to render the scene from a particular camera, we first find the world<-camera and camera<-world transforms by traversing the path from the world node to the camera node.

Qworld<-camera = C1C2...Cn
Qcamera<-world = Qworld<-camera-1 = [C1C2...Cn]-1 = Cn-1...C2-1C1-1

Then, we start at the root node of the scene graph and use the camera<-world transform as the initial modeling transform. Now, all points will be automatically transformed into the camera's frame of reference by the modeling transform.

The lookat transformation


The standard camera placement transformation is represented in GLIDE by the lookat transformation. Although the most common use for this transformation is to place a camera at one point in space looking at another point in space, it can also be used to place any object or light in the scene. The lookat transformation is particularly useful for pointing a spotlight at something in the scene.

The lookat transformation specifies a new coordinate system, (u,v,n) within the current system by

  • the origin of the new system (eye),
  • a point that will lie on the -n-axis (vrp),
  • and an up vector that should project onto the v-axis in the uv-plane (up).

In order to compute the transformation matrix for the lookat transform, we must first find the orthonormal coordinate axes (u,v,n) and the origin c.

  • c = eye
  • n = c - vrp / |c - vrp|
  • u = up x n / |up x n|
  • v = n x u
If the up vector is parallel to the vector between the eye and the vrp, then the u and v axes will be degenerate. In this case, do whatever you like with u and v, so long as (u,v,n) is orthonormal, since the operation is not well defined.

Now that we have the necessary information: the origin, c, and the coordinate axes, (u,v,n), we can compute the lookat transformation, Qlookat. The matrix, Qlookat, will transform points in the camera's or object's system into the world system just like any other modeling transform.

Let's call puvn=(pu pv pn) and pxyz=(px py pz)

pxyz=Qlookatpuvn
pxyz= c + (puu + pvv + pnn)

pxyz = T(c) Qxyz<-uvn puvn = T(c)

ux
vx
nx
0

uy
vy
ny
0

uz
vz
nz
0

0
0
0
1

pu

pv

pn

1

Note that Qxyz<-uvn is an orthogonal matrix (Qxyz<-uvn-1=Qxyz<-uvnT). So it is easy to find Quvn<-xyz=Qxyz<-uvn-1.

so we get Qlookat = T(c) Qxyz<-uvn where Qxyz<-uvn =

ux
vx
nx
0

uy
vy
ny
0

uz
vz
nz
0

0
0
0
1

and Quvn<-xyz =

ux
uy
uz
0

vx
vy
vz
0

nx
ny
nz
0

0
0
0
1

We also need to find Qlookat-1, which we can do in a straightforward fashion:

Qlookat-1 = (T(c) Qxyz<-uvn)-1 = Qxyz<-uvn-1 T(c)-1 = Qxyz<-uvnT T(c)-1 = Quvn<-xyz T(-c)

Normalizing Transform

We start out with some arbitrary view frustum or parallelepiped as defined by the frustum parameter in the GLIDE camera statement. We need to transform this volume into a canonical volume so that we can clip our polygons against standard clipping bounds and apply a standard projection to their vertices. The normalizing operation is slightly different for the parallel and perspective projections.

Parallelepiped Normalization

We start with a parallelepiped as defined by the frustum, (xmin, ymin, zmin) (xmax, ymax, zmax). The (xmin, ymin) and (xmax, ymax) pairs describe the rectangular cross-section in the z=-1 plane. The zmin and zmax values describe the back and front planes. The direction of projection (dop) is from the origin to the midpoint of that rectangle.

First we have to shear the volume so that all of the boundaries are axially aligned and the dop lies along the -z-axis. Let ym = (ymin+ymax)/2 and xm = (xmin+xmax)/2. Then our shear matrix is:

1
0
xm
0

0
1
ym
0

0
0
1
0

0
0
0
1

Next, we translate the volume so that the front clipping plane corresponds to the xy-plane. This involves just a translation by -zmax in z, T(0,0,-zmax)

Finally, we scale the volume non-uniformly to match the bounds:

-1 <= x <= 1,
-1 <= y <= 1,
-1 <= z <= 0.

with the scale transformation:
S(1/(xmax-xm), 1/(ymax-ym), 1/(zmax-zmin))

Frustum Normalization

We start with a frustum as defined by the frustum, (xmin, ymin, zmin) (xmax, ymax, zmax). The (xmin, ymin) and (xmax, ymax) pairs describe the rectangular cross-section in the z=-1 plane. The zmin and zmax values describe the back and front planes.

First we have to shear the volume so that the center of the frustum lies along the -z-axis. Let ym = (ymin+ymax)/2 and xm = (xmin+xmax)/2. Then our shear matrix is:

1
0
xm
0

0
1
ym
0

0
0
1
0

0
0
0
1

Then, we scale the volume non-uniformly to match the bounds:

z <= x <= -z,
z <= y <= -z,
-1 <= z <= a,

for some value of a<0. First we scale x and y so that the bounds of the frustum lie in the x=z, x=-z, y=z, y=-z planes: S(1/(xmax-xm), 1/(ymax-ym), 1)
Then we scale the whole frustum uniformly by -1/zmin so that the back clipping plane lies in the z=-1 plane. S(-1/zmin, -1/zmin, -1/zmin)
which gives us a value for a of a = -zmax/zmin.

Perspective Projection


Now, we are in a normalized space oriented with the camera sitting at the origin, pointing down the -z-axis. If we take any point in the scene, say p=(px,p,y,pz), we want to find the projection of that point onto the image plane, z=-1. This projection is the intersection of the ray starting at the origin and passing through the point, p.

We can find the coordinates of that point using an old trick from geometry, similar triangles. The triangle shown in dashed lines in the picture has height=py and width=-pz. The triangle formed by the projection ray, the image plane and the -z-axis has width=1. So, by similar triangles, the height of that triangle is py/(-pz).

This value, py/(-pz), is the projected y-value of the point. Similarly, the projected x-value is px/(-pz). This non-linear operation can be performed in homogeneous coordinates by allowing a w-value that is not 1. We simply apply this operation: w' = -z

Unfortunately, this operation is not invertible because the new z and w values are linearly dependent. Also, after the homogeneous normalization (division by w) has been applied, all of the z values will be -1. In order to maintain z ordering, we need to transform z to some other value. We will use the value:

z' = (z-a) / (1+a)

where a is the value of the front z clipping plane after the frustum normalization. We pick this value because this transformation will map our canonical view volume into the normalized half cube after the homogeneous division has been applied
-1 <= x <= 1,
-1 <= y <= 1,
-1 <= z <= 0.

The transformation matrix that applies this perspective transformation is:

Qproj<-vrc =

1
0
0
0

0
1
0
0

0
0
1/(1+a)
-a/(1+a)

0
0
-1
0

(as given in Foley and vanDam on p. 275) and its inverse is

Qproj<-vrc-1 =

1
0
0
0

0
1
0
0

0
0
0
-1

0
0
-(1+a)/a
-1/a

Projections in FlatLand
Visualizing the effect of a 4D transformation is pretty tough, so here is an analogous 3D transformation. The x-coordinate in our regular space is transformed just like the y-coordinate is here.

Here we have a homogeneous 2D space with no x-coordinate. The figure to the right shows the normalized frustum in the w=1 plane just before the perspective transformation (warp). The figures below show the (y,z) view volume after its perspective warp.

The transformed homogeneous plane viewed along the y-axis
The transformed homogeneous plane viewed along the z-axis
The transformed homogeneous plane viewed along the w-axis
The transformed homogeneous (y,z,w) plane




Back-Face Culling in Homogeneous Coordinates

We started out with a plane equation N.P = 0 in the object coordinate system, with N = [a b c d].

For a parallel projection: Once we are in the normalized half-space, a plane is front facing if the z-component of the normal is positive, i.e. c > 0.

For a perspective projection: Once we are in the normalized frustum, a plane is front facing (its normal points towards the camera/origin) if d > 0. After we have applied the perspective warp to the plane equation (Qproj<-vrc-1)TN, we have a new value c' = -(1+a)d/a. So, if this value is positive, (c' > 0), then we have -(1+a)d/a > 0 and thus d > 0.

Therefore, we only have to check the value for c in the final plane equation after perspective warp to determine whether a plane faces the camera and we do not have to homogenize the plane equation. Homogenizing the plane equation can reverse the direction of the normal and lead to incorrect culling.

Clipping in Homogeneous Coordinates

Once we have performed the above normalizations and the perspective projection, we have this set of clipping bounds:

-w <= x <= w
-w <= y <= w
-w <= z <= 0

This set of bounds will work for both the parallel and the perspective projection cases.

In the parallel case, w=1, so the bounds become:

-1 <= x <= 1
-1 <= y <= 1
-1 <= z <= 0

In the perspective case, w=-z, so the x and y bounds become:

z <= x <= -z
z <= y <= -z

The z bounds come from the new z value: z' = (z-a) / (1+a). At the front clipping plane, z=a, we now have z'=0 and at the back clipping plane, z=-1, we now have z'=-1 but we also have w=1 so z'=-w

Clipping against these planes is nearly as easy as clipping against the axially aligned planes. For example, suppose we are clipping two points, p1=(x1,y1,z1) and p2=(x2,y2,z2) against the x=w plane. We have already determined that the two points lie on opposite sides of the plane (i.e. for one of them x>w and for the other x).
We want to find the intersection point: pt = p1 + t(p2 - p1) with xt=wt.
This gives us the formula

x1 + t(x2 - x1)  =  w1 + t(w2 - w1)
x1 - w1  =  t[(w2 - w1)  -  (x2 - x1)]
t = (x1 - w1)/ [(w2 - w1)  -  (x2 - x1)]

Because this parameter, t is computed from coordinates that are all linearly derived from the initial, object space, coordinates of the two points, it can be used to correctly compute interpolated values along the edges of the polygon. If the points were normalized (divided through by w) before the clip, the parameter value would not be the correct value. For example, consider the midpoint in screen space of a line that starts very close to the viewer and ends very far away. The visual midpoint will be much closer to the near end of the line than to the far end of the line in object space although the homogenized parameter value would be t=0.5.

Mental Gymnastics

Consider the following scene: You are sitting on a ping pong table that is 8 feet long (in y) and 4 feet wide (in x). This table defines the xy-plane. Your eye is exactly 3 feet above the center of the table (at [0 0 3]) and you are staring directly parallel to the length of the table (along the y-axis). Use a canonical [-1,1]x[-1,1] viewing angle and an image plane one foot in front of your eye.

Questions:

  • What are the coordinates of the four corners of the ping pong table?
  • How would you describe, in GLIDE, the orientation of your eye as a camera?
  • Where are the projections of the four table points in the xy-image plane? Draw the four projected points connected in order.
  • Where are the correct projections of the edges of the table? Draw the four projected points and describe where the edges should be.
  • What does the clipped table polygon look like with a near clipping plane of z=-1 and a far clipping plane of z=-10

Last modified: Tue Mar 31 19:51:50 1998

  • 本文已收录于以下专栏:

相关文章推荐

unity3D引擎:2D游戏自动瞄准算法实现

在很多飞行射击类游戏里,都有敌人向玩家自动瞄准并开火的功能。在这里本人用unity3D引擎新版本的2D系统来实现这个功能。     首先,让我们了解一下原理的理论知识。我们可以把敌人和玩家放到一...
  • ybhjx
  • ybhjx
  • 2017年03月04日 20:07
  • 494

【转】Unity 3D 引擎中数据交互问题

第一节:必备的基础知识       在做Unity 3D引擎数据交互问题之前,需要必备以下知识。1,您对Unity 3D要进行有针对性的了解,不要求您对Unity 3D有更深入的了解,在网上找些相关的...

(转)《如何学好游戏3D引擎编程》

此篇文章献给那些为了游戏编程不怕困难的热血青年,它的神秘要我永远不间断的去挑战自我,超越自我,这样才能攀登到游戏技术的最高峰   这篇文章是我一年半前,找工作时候写的,那时是发到学校的BBS上。...

(转)如何学好3D游戏引擎编程

前言   本人不才,姑且以自己经验来谈如何学游戏编程,这个话题我酝酿了很久,它确实很难写,因为我也不是资深高手,才刚刚登堂入室,游戏编程大师功力一般都十几年左右,所以让我这样的小卒写的深入浅出更何谈容...

常用的开源3D游戏引擎(转)

2010-05-20 8:32 原文:http://hereland.com/2009-05-top-opensource-game-engine/ 开源即开放原代...

[转]Unity3D游戏引擎之构建游戏框架与导出IOS项目

转载自雨松MOMO的博客原文地址:http://blog.csdn.net/xys289187120/article/details/6862563 Unity3D 游戏引擎之构建游戏框架与...

[转]阅读一款3D引擎的方法备忘

链接:http://www.cnitblog.com/linghuye/archive/2009/11/29/62870.html 要带着引擎系统所共有的问题去阅读,不要被具体引擎的实现思路牵着...
  • arundev
  • arundev
  • 2012年06月24日 20:01
  • 285

3D游戏引擎术语介绍.doc

  • 2015年06月13日 21:28
  • 527KB
  • 下载

3D图象算法(转)

可视化编辑器,alt+1到9键,切换到工具区,tab键,选择按钮,esc键,返回编辑  3D简介    我们首先从坐标系统开始。你也许知道在2D里我们经常使用Ren?笛卡儿坐标系统在平面上来识别...
内容举报
返回顶部
收藏助手
不良信息举报
您举报文章:【转】3D引擎中的剪切算法
举报原因:
原因补充:

(最多只允许输入30个字)