拉普拉斯滤波器产生奇怪的结果（Java）

Question

我正在尝试应用

[-1][-1][-1]
[-1][8][-1]
[-1][-1][-1]

3*3 拉普拉斯滤波器到著名照片的灰度版本（png 扩展名）在这里。

我主要使用BufferedImage 类来处理图像。这是拉普拉斯滤波器方法。

private BufferedImage measureContrast(BufferedImage image) {
        BufferedImage grayScale = createGrayscaleImage(image);
        BufferedImage copy = copyImage(grayScale);

        int width = image.getWidth();
        int height = image.getHeight();
        int sum=0;
        int a;
         //3*3 Laplacian filter (-1,-1,-1), (-1,8,-1), (-1,-1,-1)
        for(int y=1;y<height-1;y++)
            for(int x=1;x<width-1;x++) {
                sum = (-1*(grayScale.getRGB(x-1, y-1)&0xff)) + (-1*(grayScale.getRGB(x, y-1)&0xff)) + (-1*(grayScale.getRGB(x+1, y-1)&0xff))
                        + (-1*(grayScale.getRGB(x-1, y)&0xff)) + (8*(grayScale.getRGB(x,y)&0xff)) + (-1*(grayScale.getRGB(x+1, y)&0xff)) +
                        (-1*(grayScale.getRGB(x-1, y+1)&0xff)) + (-1*(grayScale.getRGB(x, y+1)&0xff)) + (-1*(grayScale.getRGB(x+1, y+1)&0xff));             
                a = ((grayScale.getRGB(x, y)>>24)&0xff);
                copy.setRGB(x, y, ((a<<24)|(sum<<16)|(sum<<8)|(sum)));
            }
return copy;

如果我运行该代码，结果是这样的

这显然是错误的。图像中突然出现了奇怪的粗线。

我可以保证灰度版本是正确的，因为当我在应用过滤器之前运行以下代码时，输​​出会给出完美的灰度图像。

private BufferedImage measureContrast(BufferedImage image) {
        BufferedImage grayScale = createGrayscaleImage(image);
        BufferedImage copy = copyImage(grayScale); /*rest of the code is commented*/
return copy;

我已经尝试找到问题几个小时了，但我认为代码没有任何问题...任何见解将不胜感激。提前致谢！

---

为了复制图像，我使用了以下代码。我从堆栈溢出中选择的答案中借用了它，所以我认为它不会错。

BufferedImage copyImage(BufferedImage bi) {
         ColorModel cm = bi.getColorModel();
         boolean isAlphaPremultiplied = cm.isAlphaPremultiplied();
         WritableRaster raster = bi.copyData(null);
         return new BufferedImage(cm, raster, isAlphaPremultiplied, null);
    }

另外，结果图像是这样打印的

BufferedImage Contrast = measureContrast(image);
//write image
try {
    ImageIO.write(Contrast, "png", new File(outputPath));
    System.out.println("Printing complete");
}catch(IOException e) {
    System.out.println("File Printing Error: "+e);
}

以防万一，这里是灰度图像生成方法。

private BufferedImage createGrayscaleImage(BufferedImage image) {
        int width = image.getWidth();
        int height = image.getHeight();
        BufferedImage copy = copyImage(image);

        int p=0, a=0, r=0, g=0, b=0, avg=0;
        for(int y=0;y<height;y++)
            for(int x=0;x<width;x++) {
                p=image.getRGB(x, y);
                a=(p>>24)&0xff;
                r=(p>>16)&0xff;
                g=(p>>8)&0xff;
                b=p&0xff;
                avg = (r+g+b)/3;
                p = (a<<24) | (avg<<16) | (avg<<8) | avg;
                copy.setRGB(x, y, p);
            }
        return copy;
    }

Answer 1

private BufferedImage measureContrast(BufferedImage image) {
        BufferedImage grayScale = createGrayscaleImage(image);
        BufferedImage copy = copyImage(grayScale);

        int width = image.getWidth();
        int height = image.getHeight();
        int sum=0;
        int a;
         //3*3 Laplacian filter (-1,-1,-1), (-1,8,-1), (-1,-1,-1)
        for(int y=1;y<height-1;y++)
            for(int x=1;x<width-1;x++) {
                sum = (-1*(grayScale.getRGB(x-1, y-1)&0xff)) + (-1*(grayScale.getRGB(x, y-1)&0xff)) + (-1*(grayScale.getRGB(x+1, y-1)&0xff))
                        + (-1*(grayScale.getRGB(x-1, y)&0xff)) + (8*(grayScale.getRGB(x,y)&0xff)) + (-1*(grayScale.getRGB(x+1, y)&0xff)) +
                        (-1*(grayScale.getRGB(x-1, y+1)&0xff)) + (-1*(grayScale.getRGB(x, y+1)&0xff)) + (-1*(grayScale.getRGB(x+1, y+1)&0xff));             
                a = ((grayScale.getRGB(x, y)>>24)&0xff);
                copy.setRGB(x, y, ((a<<24)|(sum<<16)|(sum<<8)|(sum)));
            }
return copy;

这个方法是错误的。当 sum 为负值时，我应该考虑过。所以我在嵌套的 for 循环中添加了sum = (sum>0)?sum:0;，以确保在这些情况下像素值为 0。