Camera3D

Set up view and projection matrices for 3D visualizations — point clouds, surface meshes, and volumetric data.

API

class Camera3D {
public:
    Camera3D();

    void lookAt(const Eigen::Vector3f& eye,
                const Eigen::Vector3f& target,
                const Eigen::Vector3f& up = Eigen::Vector3f::UnitY());

    void setPerspective(float fovDegrees, float aspect,
                        float nearPlane, float farPlane);

    auto viewMatrix()           const -> Eigen::Matrix4f;
    auto projectionMatrix()     const -> Eigen::Matrix4f;
    auto viewProjectionMatrix() const -> Eigen::Matrix4f;

    auto position() const -> const Eigen::Vector3f&;
    auto target()   const -> const Eigen::Vector3f&;
};

Methods

Method	Description
lookAt(eye, target, up)	Position camera at eye, looking at target
setPerspective(fov, aspect, near, far)	Set perspective projection parameters
viewMatrix()	Returns the 4×4 view (camera) matrix
projectionMatrix()	Returns the 4×4 perspective projection matrix
viewProjectionMatrix()	Returns projection × view

Default Values

Parameter	Default
Eye position	(0, 0, 5)
Target	(0, 0, 0)
Up	(0, 1, 0)
FOV	45°
Aspect	1.0
Near	0.1
Far	100.0

View Matrix

The view matrix transforms world coordinates to camera space using a standard right-handed look-at construction:

$$V = \begin{bmatrix} r_x & r_y & r_z & -\mathbf{r} \cdot \mathbf{e} \\ u_x & u_y & u_z & -\mathbf{u} \cdot \mathbf{e} \\ -f_x & -f_y & -f_z & \mathbf{f} \cdot \mathbf{e} \\ 0 & 0 & 0 & 1 \end{bmatrix}$$

where $\mathbf{f} = \text{normalize}(\text{target} - \text{eye})$, $\mathbf{r} = \mathbf{f} \times \mathbf{up}$, $\mathbf{u} = \mathbf{r} \times \mathbf{f}$.

Example

Skigen::Plot::Camera3D camera;
camera.lookAt({3.f, 3.f, 3.f}, {0.f, 0.f, 0.f});
camera.setPerspective(45.f, 16.f / 9.f, 0.1f, 50.f);

Skigen::Plot::PlotView view;
view.pointCloud(vertices);
view.setCamera(camera);
view.show();