Question

我正在编写raytracer，并想要实现一个用户界面。当我用它初始化渲染而不是通过main进行渲染时，会得到不同的结果。

我尝试过 +禁用多线程， +不建议在单独的线程中启动GUI +我正在使用（就我而言）完全相同的代码，无论是否是GUI

//in Main-class
public static void main(String[] args){

        Window renderWindow = new Window(IMAGE_WIDTH, IMAGE_HEIGHT, "Blank");
        draw(renderWindow);

        //Start UserInterface
//        UserInterface userInterface = new UserInterface();
//        userInterface.setVisible(true);

//        SwingUtilities.invokeLater(() -> {
//            UserInterface userInterface = new UserInterface();
//            userInterface.setVisible(true);
//        });
    }

private static void draw(Window renderWindow){
        Scene renderScene = new Scene();

        setupScene(renderScene);

        raytraceScene(renderWindow, renderScene);
}
private static void raytraceScene(Window renderWindow, Scene renderScene){

        Raytracer raytracer = new Raytracer(
                renderScene,
                renderWindow,
                1, //number of recursions
                2, //anti-aliasing factor
                false); //use multithreading yes/no

        raytracer.renderScene();
    }

在setupScene（renderScene）灯光，对象，cornellbox和照相机中进行设置。

//in UserInterface-class
public UserInterface() {

        add(root);
        setTitle("Make Raytracing great again!");
        setSize(600, 400);
        setDefaultCloseOperation(JFrame.DISPOSE_ON_CLOSE);

        shapelistModel = new DefaultListModel<>();
        list_shapes = new JList<>(shapelistModel);

        renderScene = new Scene();

        renderButton.addActionListener(evt -> {
            //Set the title for the rendered image
            String renderTitle = textField_Title.getText();

            //Create a window to draw the image in
            Window renderWindow = new Window(IMAGE_WIDTH, IMAGE_HEIGHT, renderTitle);

            //Do the magic
            draw(renderWindow);
        });
    }

这里的draw-Method与Main-class中的相同；除此之外，按一下按钮不会覆盖renderScene变量，因为我已经在构造函数中创建了一个场景

换句话说，是否有一种方法可以将图像的渲染拆分到系统中的所有核心，具体取决于有多少核心？我已经尝试通过Runtime.getRuntime（）。availableProcessors（）获取内核数，并将图像拆分为多个块，但是似乎无法正常工作。也许我对每个线程的x和y的起点和终点的计算是错误的？

int coreCount = Runtime.getRuntime().availableProcessors();
            int halfCoreCount = coreCount/2;

            int imageWidth = mBufferedImage.getWidth();
            int imageHeight = mBufferedImage.getHeight();
//
            int imageWidthByCoreCount = imageWidth/halfCoreCount;
            int imageHeightByCoreCount = imageHeight/halfCoreCount;

            //Get an array of workers with the length of the available cores on this machine
            Thread[] renderFrames = new Thread[coreCount];

            //Create the workers, assign them their boxes and let them render
            for (int i = 0; i < renderFrames.length; i++) {

                /* Calculate start and end-point for x and y boundaries for each Thread */
                //Use modulo to loop back to the beginning when the row/column of Threads was completed
                int xStart = (imageWidthByCoreCount * i); // % halfCoreCount;
                int xEnd = (imageWidthByCoreCount * (i + 1)); // % halfCoreCount;

                int yStart = (imageHeightByCoreCount * i); // % halfCoreCount;
                int yEnd = (imageHeightByCoreCount * (i + 1)); // % halfCoreCount;

                Log.print(this, "x: " + xStart + "; " + xEnd + " | y: " + yStart + "; " + yEnd);

                //Assign new worker to the Thread
                renderFrames[i] = new Thread(new Worker(this.mBufferedImage, xStart, xEnd, yStart, yEnd));

                //Get the worker going
                renderFrames[i].start();
            }

取模没有奏效，所以我尝试了一下，但无济于事。

That's how it looks when I'm using the GUI

And that's how it's (for now) supposed to look

即使使用相同的代码，使用GUI时程序的呈现方式也会有所不同？

0 个答案: