Box2D C++ 教程-射线投射

声明：本教程翻译自：Box2D C++ tutorials - Ray casting，仅供学习参考。

射线投射经常用来找出部分物理世界当中已经存在的物件。所谓射线就是一条直线，我们可以通过使用Box2D当中提供的方法来检测是否与某个定制器（fixture）有交点。我们也可以找出与定制器交点的法线。

下面就是上面我提到的方法，如果射线与某个定制器发生碰撞则返回true。这里需要注意的是此方法是b2Fixture类的一个成员，这也就是说首先我们满足一个射线对儿这个基本条件才行。

1	bool b2Fixture::RayCast(b2RayCastOutput* output, const b2RayCastInput& input);

现在让我们看一下输入、输出参数。直接从源代码中可以看出b2RayCastInput包含了哪些参数：

1 2 3 4 5 6

// Ray-cast input data. The ray extends from p1 to p1 + maxFraction * (p2 - p1). struct b2RayCastInput { b2Vec2 p1, p2; float32 maxFraction; };

p1和p2点确定了射线的方向。maxFraction表示了射线检测到交点的距离。下面的图可能能够让人一目了然。maxFraction等于1意味着射线段是从p1点到p2点，在此例子中显然不能与多边形有交点，但是maxFraction等于2就可以。

下面是b2RayCastOutput包括的参数：

1 2 3 4 5 6 7

// Ray-cast output data. The ray hits at p1 + fraction * (p2 - p1), where p1 and p2 // come from b2RayCastInput. struct b2RayCastOutput { b2Vec2 normal; float32 fraction; };

如果射线确实与多边形相交，b2Fixture::RayCast将会返回true，而且我们可以通过这个输出结构体里的fraction参数找出交点段值（详见上图），还有在多边形定制器表面生成的法线：

例子

为了尝试一下这个很方便实用的功能，让我们创建一个场景，场景的条件为失重状态，其中漂浮着一些多边形。现在创建这些你应该很顺手了，接下来我们在四周创建一些墙来代替四方盒子作为边界，为的是看起来更好看。

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

FooTest() { //a static body b2BodyDef myBodyDef; myBodyDef.type = b2_staticBody; myBodyDef.position.Set(0, 0); b2Body* staticBody = m_world->CreateBody(&myBodyDef); //shape definition b2PolygonShape polygonShape; //fixture definition b2FixtureDef myFixtureDef; myFixtureDef.shape = &polygonShape; //add four walls to the static body b2Vec2 bl(-20, 0); b2Vec2 br( 20, 0); b2Vec2 tl(-20,40); b2Vec2 tr( 20,40); polygonShape.SetAsEdge( bl, br ); //ground staticBody->CreateFixture(&myFixtureDef); polygonShape.SetAsEdge( tl, tr);//ceiling staticBody->CreateFixture(&myFixtureDef); polygonShape.SetAsEdge( bl, tl );//left wall staticBody->CreateFixture(&myFixtureDef); polygonShape.SetAsEdge( br, tr );//right wall staticBody->CreateFixture(&myFixtureDef); myBodyDef.type = b2_dynamicBody; myBodyDef.position.Set(0,20); polygonShape.SetAsBox(2,2); myFixtureDef.density = 1; for (int i = 0; i < 5; i++) m_world->CreateBody(&myBodyDef)->CreateFixture(&myFixtureDef); //circles b2CircleShape circleShape; circleShape.m_radius = 2; myFixtureDef.shape = &circleShape; for (int i = 0; i < 5; i++) m_world->CreateBody(&myBodyDef)->CreateFixture(&myFixtureDef); //turn gravity off m_world->SetGravity( b2Vec2(0,0) ); }

现在我们需要有一个射线与这些多边形发生碰撞。让我们在屏幕的中心生成一条四周投射的射线，并且缓慢旋转。我们需要唯一做的就是保持住射线的旋转角度，那么我们就不要为此特别封装一个类了，我们只需要有一个全局变量就OK了。

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

//at global scope float currentRayAngle = 0; //in Step() function currentRayAngle += 360 / 20.0 / 60.0 * DEGTORAD; //one revolution every 20 seconds //calculate points of ray float rayLength = 25; //long enough to hit the walls b2Vec2 p1( 0, 20 ); //center of scene b2Vec2 p2 = p1 + rayLength * b2Vec2( sinf(currentRayAngle), cosf(currentRayAngle) ); //draw a line glColor3f(1,1,1); //white glBegin(GL_LINES); glVertex2f( p1.x, p1.y ); glVertex2f( p2.x, p2.y ); glEnd();

现在在场景中你应该能够看到一条旋转的白线。现在我们可以就用RayCast方法来获取离射线最近的多边形，并画出线段的长度。我们将会对每个多边形中的每个定制器进行检测，当然这并不是最有效的办法，但是可以作为一个学习的例子（详见查询world话题）来学习。这样的话，我们就可以了解一下世界里的物件：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

//in Step() function, continuing on from section above //set up input b2RayCastInput input; input.p1 = p1; input.p2 = p2; input.maxFraction = 1; //check every fixture of every body to find closest float closestFraction = 1; //start with end of line as p2 b2Vec2 intersectionNormal(0,0); for (b2Body* b = m_world->GetBodyList(); b; b = b->GetNext()) { for (b2Fixture* f = b->GetFixtureList(); f; f = f->GetNext()) { b2RayCastOutput output; if ( ! f->RayCast( &output, input ) ) continue; if ( output.fraction < closestFraction ) { closestFraction = output.fraction; intersectionNormal = output.normal; } } } b2Vec2 intersectionPoint = p1 + closestFraction * (p2 - p1); //draw a line glColor3f(1,1,1); //white glBegin(GL_LINES); glVertex2f( p1.x, p1.y ); glVertex2f( intersectionPoint.x, intersectionPoint.y ); glEnd(); //draw a point at the intersection point glPointSize(5); glBegin(GL_POINTS); glVertex2f( intersectionPoint.x, intersectionPoint.y ); glEnd();

你可能注意到现在我们在每个上方画出了两条线…为了清晰起见，删除第一条你就会看到类似下面的画面：

上面所说的就是找出交点。我们也可以在输出结构中使用法线来实现一些有趣的想法，现在我们就可以试试看。首先，让我们把法线长什么样儿简单的画出来：

1 2 3 4 5 6 7

//draw intersection normal b2Vec2 normalEnd = intersectionPoint + intersectionNormal; glColor3f(0,1,0); //green glBegin(GL_LINES); glVertex2f( intersectionPoint.x, intersectionPoint.y ); glVertex2f( normalEnd.x, normalEnd.y ); glEnd();

作为最后的压轴戏，我们可以把投射代码放到一个自有方法中，然后每次不断的递归调用直到射线的耗尽。这并不意味着Box2D怎么写都行，我只是觉得这么做代码看起来更整洁而已:)。

代码：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

//new function for FooTest class void drawReflectedRay( b2Vec2 p1, b2Vec2 p2 ) { //set up input b2RayCastInput input; input.p1 = p1; input.p2 = p2; input.maxFraction = 1; //check every fixture of every body to find closest float closestFraction = 1; //start with end of line as p2 b2Vec2 intersectionNormal(0,0); for (b2Body* b = m_world->GetBodyList(); b; b = b->GetNext()) { for (b2Fixture* f = b->GetFixtureList(); f; f = f->GetNext()) { b2RayCastOutput output; if ( ! f->RayCast( &output, input ) ) continue; if ( output.fraction < closestFraction ) { closestFraction = output.fraction; intersectionNormal = output.normal; } } } b2Vec2 intersectionPoint = p1 + closestFraction * (p2 - p1); //draw this part of the ray glBegin(GL_LINES); glVertex2f( p1.x, p1.y ); glVertex2f( intersectionPoint.x, intersectionPoint.y ); glEnd(); if ( closestFraction == 1 ) return; //ray hit nothing so we can finish here if ( closestFraction == 0 ) return; //still some ray left to reflect b2Vec2 remainingRay = (p2 - intersectionPoint); b2Vec2 projectedOntoNormal = b2Dot(remainingRay, intersectionNormal) * intersectionNormal; b2Vec2 nextp2 = p2 - 2 * projectedOntoNormal; //recurse drawReflectedRay(intersectionPoint, nextp2); }

…然后在Step()方法中天假下面几行代码：

1 2 3 4 5 6 7

//calculate points of ray float rayLength = 25; b2Vec2 p1( 0, 20 ); //center of scene b2Vec2 p2 = p1 + rayLength * b2Vec2( sinf(currentRayAngle), cosf(currentRayAngle) ); glColor3f(1,1,1); //white drawReflectedRay(p1, p2);