Question

我正在编写小型raytracer，我想围绕某些物体旋转相机。这样我就可以实现以下代码：

vecTemp是相机位置向量，vecT是第一个位置（我在for循环中运行此代码段）。偏航角是移动摄像机的弧度值角。

double yaw = degree * (PI / 180.f);
vecTemp.x() = cos(yaw) * vecT.x() - sin(yaw) * vecT.z();
vecTemp.z() = cos(yaw) * vecT.z() + sin(yaw) * vecT.x();
vecTemp.y() = 0;
sett.camera.buildCameraToWorld(vecTemp, { 0, 0, 0} );
sett.camera.rotateY(2*degree);

void buildCameraToWorld(const arma::dvec3& from, const arma::dvec3& to, const arma::dvec3& tmp = arma::dvec3{ 0, 1, 0 })
   {
      arma::dvec3 forward = arma::normalise(from - to);
      arma::dvec3 right = arma::cross(arma::normalise(tmp), forward);
      arma::dvec3 up = arma::cross(forward, right);

      this->cameraToWorld.zeros();

      this->cameraToWorld(0, 0) = right.x();
      this->cameraToWorld(0, 1) = right.y();
      this->cameraToWorld(0, 2) = right.z();
      this->cameraToWorld(1, 0) = up.x();
      this->cameraToWorld(1, 1) = up.y();
      this->cameraToWorld(1, 2) = up.z();
      this->cameraToWorld(2, 0) = forward.x();
      this->cameraToWorld(2, 1) = forward.y();
      this->cameraToWorld(2, 2) = forward.z();

      this->cameraToWorld(3, 0) = from.x();
      this->cameraToWorld(3, 1) = from.y();
      this->cameraToWorld(3, 2) = from.z();

      this->cameraToWorld = this->cameraToWorld.t();
   }
void rotateY(double yaw)
   {
      yaw = yaw * (PI / 180);
      this->cameraToWorld = this->cameraToWorld * arma::dmat44{ { cos(yaw), 0, sin(yaw), 0},
                                                                {        0, 1,        0, 0},
                                                                {-sin(yaw), 0, cos(yaw), 0},
                                                                {        0, 0,        0, 1} };
   }

所以，现在我不明白，为什么平移后我会在Y轴上旋转-在buildCameraToWorld之后，我应该得到一个新的相机矩阵来查看我的原点，对吗？为什么我需要旋转双度值？

我发现了另一种实现方式-我可以旋转整个世界并且不像在OpenGL的相机中那样移动相机，但是对于光线追踪器来说这是个好主意吗？

旋转光线追踪器相机

0 个答案: