VTK 的可视化方法：流线、流管、流面、流带

VTK 的可视化方法：流线、流管、流面、流带

2维流线示例：

本文章主要讲解3维的流线、流管、流面、流带的构造方法。

三种相似的可视化方法

1. 流线（Streamlines）：每个点速度切线方向连成的线。
2. 迹线（Pathlines）：粒子实际的轨迹线。
3. 脉线（Streaklines）：连续时刻出发的粒子在某一时刻的连线。

流线生成使用的类

• vtkRungeKutta4：四阶龙格库塔 (Runge-Kutta) 求解微分。
• vtkStreamTracer：通过整合矢量场生成流线。

实例：单一流线

完整代码：

#include "VTKStreamline.h"

#include <vtkConeSource.h>
#include <vtkMultiBlockPLOT3DReader.h>
#include <vtkDataSet.h>
#include <vtkMultiBlockDataSet.h>
#include <vtkRungeKutta4.h>
#include <vtkStreamTracer.h>
#include <vtkDataArray.h>
#include <vtkPointData.h>
#include <vtkShrinkPolyData.h>
#include <vtkStructuredGridGeometryFilter.h>
#include <vtkStructuredGridOutlineFilter.h>
#include <vtkContourFilter.h>
#include <vtkPolyDataMapper.h>
#include <vtkActor.h>
#include <vtkRenderer.h>
#include <vtkRenderWindow.h>

VTKStreamline::VTKStreamline(QWidget* parent)
	: QMainWindow(parent)
{
	ui.setupUi(this);

	_pVTKWidget = new QVTKOpenGLNativeWidget();
	this->setCentralWidget(_pVTKWidget);
	// this->showMaximized();

	// 1. generate data
	// vtkSmartPointer<vtkConeSource> cone = vtkSmartPointer<vtkConeSource>::New();
	// or, read data
	// vtkMultiBlockPLOT3DReader 是一个读取器对象，用于读取 PLOT3D 格式的文件并在输出时生成结构化网格
	vtkSmartPointer<vtkMultiBlockPLOT3DReader> plot3dReader = vtkSmartPointer<vtkMultiBlockPLOT3DReader>::New();
	plot3dReader->SetXYZFileName("combxyz.bin");
	plot3dReader->SetQFileName("combq.bin");
	plot3dReader->SetScalarFunctionNumber(100);
	plot3dReader->SetVectorFunctionNumber(202);
	qDebug() << plot3dReader->GetOutput()->GetNumberOfBlocks(); // 0
	// 反向更新管线
	plot3dReader->Update();
	qDebug() << plot3dReader->GetOutput()->GetNumberOfBlocks(); // 1
	vtkDataSet* plot3dOutput = (vtkDataSet*)(plot3dReader->GetOutput()->GetBlock(0));
	// 四阶龙格库塔 (Runge-Kutta) 求解微分
	vtkSmartPointer<vtkRungeKutta4> integ = vtkSmartPointer<vtkRungeKutta4>::New();
	// 通过整合矢量场生成流线
	vtkSmartPointer<vtkStreamTracer> streamer = vtkSmartPointer<vtkStreamTracer>::New();
	streamer->SetIntegrator(integ);
	streamer->SetInputData(plot3dOutput);
	streamer->SetStartPosition(15, 5, 32);
	streamer->SetMaximumPropagation(100);
	streamer->SetInitialIntegreationStep(0.1);
	streamer->SetIntegreationDirectionToBackward();

	// 2. filter
	// 产生结构化栅格边界的一个线轮廓
	vtkSmartPointer<vtkStructuredGridOutlineFilter> outline = vtkSmartPointer<vtkStructuredGridOutlineFilter>::New();
	outline->SetInputData(plot3dOutput);

	// 3. mapper
	vtkSmartPointer<vtkPolyDataMapper> outlineMapper = vtkSmartPointer<vtkPolyDataMapper>::New();
	vtkSmartPointer<vtkPolyDataMapper> singleMapper = vtkSmartPointer<vtkPolyDataMapper>::New();
	singleMapper->SetScalarRange(plot3dOutput->GetPointData()->GetScalars()->GetRange());
	// 4. actor
	vtkSmartPointer<vtkActor> outlineActor = vtkSmartPointer<vtkActor>::New();
	vtkSmartPointer<vtkActor> singleActor = vtkSmartPointer<vtkActor>::New();

	// 5. renderer
	vtkSmartPointer<vtkRenderer> renderer = vtkSmartPointer<vtkRenderer>::New();
	renderer->SetBackground(0.3, 0.6, 0.3); // Background Color: Green

	// 6. connect
	outlineMapper->SetInputConnection(outline->GetOutputPort());
	singleMapper->SetInputConnection(streamer->GetOutputPort());
	outlineActor->SetMapper(outlineMapper);
	singleActor->SetMapper(singleMapper);
	renderer->AddActor(outlineActor);
	renderer->AddActor(singleActor);

	this->_pVTKWidget->renderWindow()->AddRenderer(renderer);
	this->_pVTKWidget->renderWindow()->Render();
}

VTKStreamline::~VTKStreamline()
{}

运行效果：

实例：流管

我们再加上一个过滤器，把流线变成流管：

	vtkSmartPointer<vtkTubeFilter> streamTube = vtkSmartPointer<vtkTubeFilter>::New();
	streamTube->SetInputConnection(streamer->GetOutputPort());
	streamTube->SetRadius(0.06);
	streamTube->SetNumberOfSides(12);

流管实际上是用一个圆柱面包裹住流线，流线依旧存在。这样的显示效果会更好一点：

实例：多条流线

在之前的代码中，新增：

	vtkSmartPointer<vtkLineSource> seeds = vtkSmartPointer<vtkLineSource>::New();
	// 设置线段的端点
	seeds->SetPoint1(15, -5, 32);
	seeds->SetPoint2(15, 5, 32);
	seeds->SetResolution(21);	

	vtkSmartPointer<vtkStreamTracer> streamer2 = vtkSmartPointer<vtkStreamTracer>::New();
	streamer2->SetIntegrator(integ);
	streamer2->SetInputData(plot3dOutput);
	// streamer2->SetStartPosition(15, 5, 32);
	streamer2->SetMaximumPropagation(100);
	streamer2->SetInitialIntegreationStep(0.1);
	streamer2->SetIntegreationDirectionToBackward();
	streamer2->SetSourceConnection(seeds->GetOutputPort());
	
	vtkSmartPointer<vtkPolyDataMapper> multipleMapper = vtkSmartPointer<vtkPolyDataMapper>::New();
	multipleMapper->SetScalarRange(plot3dOutput->GetPointData()->GetScalars()->GetRange());
	vtkSmartPointer<vtkActor> multipleActor = vtkSmartPointer<vtkActor>::New();
	multipleMapper->SetInputConnection(streamer2->GetOutputPort());
	multipleActor->SetMapper(multipleMapper);
	renderer->AddActor(multipleActor);

实例：流面

新增一个过滤器 vtkRuledSurfaceFilter，把多条流线合并成一个流面：

	vtkSmartPointer<vtkRuledSurfaceFilter> scalarSurface = vtkSmartPointer<vtkRuledSurfaceFilter>::New();
	scalarSurface->SetInputConnection(streamer2->GetOutputPort());
	// 设置生成方法
	scalarSurface->SetRuledModeToPointWalk();

	// multipleMapper->SetInputConnection(streamer2->GetOutputPort());
	multipleMapper->SetInputConnection(scalarSurface->GetOutputPort());

运行结果：

实例：流带

流带其实是按每条流线拓展而成的一条条带状的面，比起流面，更能便于展示走势。

我们只需要在前面的代码中新增一行代码：

	vtkSmartPointer<vtkRuledSurfaceFilter> scalarSurface = vtkSmartPointer<vtkRuledSurfaceFilter>::New();
	scalarSurface->SetInputConnection(streamer2->GetOutputPort());
	// 设置生成方法
	scalarSurface->SetRuledModeToPointWalk();
	scalarSurface->SetOnRatio(2); // 新增代码

运行结果：

完整代码

#include "VTKStreamline.h"

#include <vtkConeSource.h>
#include <vtkLineSource.h>
#include <vtkMultiBlockPLOT3DReader.h>
#include <vtkDataSet.h>
#include <vtkMultiBlockDataSet.h>
#include <vtkRungeKutta4.h>
#include <vtkStreamTracer.h>
#include <vtkTubeFilter.h>
#include <vtkDataArray.h>
#include <vtkPointData.h>
#include <vtkShrinkPolyData.h>
#include <vtkStructuredGridGeometryFilter.h>
#include <vtkStructuredGridOutlineFilter.h>
#include <vtkContourFilter.h>
#include <vtkRuledSurfaceFilter.h>
#include <vtkPolyDataMapper.h>
#include <vtkActor.h>
#include <vtkRenderer.h>
#include <vtkRenderWindow.h>

VTKStreamline::VTKStreamline(QWidget* parent)
	: QMainWindow(parent)
{
	ui.setupUi(this);

	_pVTKWidget = new QVTKOpenGLNativeWidget();
	this->setCentralWidget(_pVTKWidget);
	// this->showMaximized();

	// 1. generate data
	vtkSmartPointer<vtkLineSource> seeds = vtkSmartPointer<vtkLineSource>::New();
	// 设置线段的端点
	seeds->SetPoint1(15, -5, 32);
	seeds->SetPoint2(15, 5, 32);
	seeds->SetResolution(21);

	// or, read data
	// vtkMultiBlockPLOT3DReader 是一个读取器对象，用于读取 PLOT3D 格式的文件并在输出时生成结构化网格
	vtkSmartPointer<vtkMultiBlockPLOT3DReader> plot3dReader = vtkSmartPointer<vtkMultiBlockPLOT3DReader>::New();
	plot3dReader->SetXYZFileName("combxyz.bin");
	plot3dReader->SetQFileName("combq.bin");
	plot3dReader->SetScalarFunctionNumber(100);
	plot3dReader->SetVectorFunctionNumber(202);
	qDebug() << plot3dReader->GetOutput()->GetNumberOfBlocks(); // 0
	// 反向更新管线
	plot3dReader->Update();
	qDebug() << plot3dReader->GetOutput()->GetNumberOfBlocks(); // 1
	vtkDataSet* plot3dOutput = (vtkDataSet*)(plot3dReader->GetOutput()->GetBlock(0));
	// 四阶龙格库塔 (Runge-Kutta) 求解微分
	vtkSmartPointer<vtkRungeKutta4> integ = vtkSmartPointer<vtkRungeKutta4>::New();
	// 通过整合矢量场生成流线
	vtkSmartPointer<vtkStreamTracer> streamer = vtkSmartPointer<vtkStreamTracer>::New();
	streamer->SetIntegrator(integ);
	streamer->SetInputData(plot3dOutput);
	streamer->SetStartPosition(15, 5, 32);
	streamer->SetMaximumPropagation(100);
	streamer->SetInitialIntegreationStep(0.1);
	streamer->SetIntegreationDirectionToBackward();
	vtkSmartPointer<vtkStreamTracer> streamer2 = vtkSmartPointer<vtkStreamTracer>::New();
	streamer2->SetIntegrator(integ);
	streamer2->SetInputData(plot3dOutput);
	// streamer2->SetStartPosition(15, 5, 32);
	streamer2->SetMaximumPropagation(100);
	streamer2->SetInitialIntegreationStep(0.1);
	streamer2->SetIntegreationDirectionToBackward();
	streamer2->SetSourceConnection(seeds->GetOutputPort());

	// 2. filter
	// 产生结构化栅格边界的一个线轮廓
	vtkSmartPointer<vtkStructuredGridOutlineFilter> outline = vtkSmartPointer<vtkStructuredGridOutlineFilter>::New();
	outline->SetInputData(plot3dOutput);
	vtkSmartPointer<vtkTubeFilter> streamTube = vtkSmartPointer<vtkTubeFilter>::New();
	streamTube->SetInputConnection(streamer->GetOutputPort());
	streamTube->SetRadius(0.06);
	streamTube->SetNumberOfSides(12);
	vtkSmartPointer<vtkRuledSurfaceFilter> scalarSurface = vtkSmartPointer<vtkRuledSurfaceFilter>::New();
	scalarSurface->SetInputConnection(streamer2->GetOutputPort());
	// 设置生成方法
	scalarSurface->SetRuledModeToPointWalk();
	scalarSurface->SetOnRatio(2);

	// 3. mapper
	vtkSmartPointer<vtkPolyDataMapper> outlineMapper = vtkSmartPointer<vtkPolyDataMapper>::New();
	vtkSmartPointer<vtkPolyDataMapper> singleMapper = vtkSmartPointer<vtkPolyDataMapper>::New();
	singleMapper->SetScalarRange(plot3dOutput->GetPointData()->GetScalars()->GetRange());
	vtkSmartPointer<vtkPolyDataMapper> multipleMapper = vtkSmartPointer<vtkPolyDataMapper>::New();
	multipleMapper->SetScalarRange(plot3dOutput->GetPointData()->GetScalars()->GetRange());

	// 4. actor
	vtkSmartPointer<vtkActor> outlineActor = vtkSmartPointer<vtkActor>::New();
	vtkSmartPointer<vtkActor> singleActor = vtkSmartPointer<vtkActor>::New();
	vtkSmartPointer<vtkActor> multipleActor = vtkSmartPointer<vtkActor>::New();

	// 5. renderer
	vtkSmartPointer<vtkRenderer> renderer = vtkSmartPointer<vtkRenderer>::New();
	renderer->SetBackground(0.3, 0.6, 0.3); // Background Color: Green

	// 6. connect
	outlineMapper->SetInputConnection(outline->GetOutputPort());
	singleMapper->SetInputConnection(streamTube->GetOutputPort());
	// multipleMapper->SetInputConnection(streamer2->GetOutputPort());
	multipleMapper->SetInputConnection(scalarSurface->GetOutputPort());
	outlineActor->SetMapper(outlineMapper);
	singleActor->SetMapper(singleMapper);
	multipleActor->SetMapper(multipleMapper);
	renderer->AddActor(outlineActor);
	renderer->AddActor(singleActor);
	renderer->AddActor(multipleActor);

	this->_pVTKWidget->renderWindow()->AddRenderer(renderer);
	this->_pVTKWidget->renderWindow()->Render();
}

VTKStreamline::~VTKStreamline()
{}