Java 查找图像二值对象的周长

我试图找到二元物体的周长

考虑下面的图片

[ 0 0 0 0 1 1 0 ]
[ 0 0 1 0 0 0 0 ]
[ 0 1 1 0 1 1 0 ]
[ 0 1 1 0 0 1 0 ]
[ 0 1 1 1 0 0 0 ]
[ 0 1 0 0 1 1 0 ]
[ 0 0 0 0 1 1 0 ]

标记的图像将如下所示

[ 0 0 0 0 1 1 0 ]
[ 0 0 2 0 0 0 0 ]
[ 0 2 2 0 3 3 0 ]
[ 0 2 2 0 0 3 0 ]
[ 0 2 2 0 0 0 0 ]
[ 0 2 0 0 4 4 0 ]
[ 0 0 0 0 4 4 0 ]

我还收集了数组列表中的每个对象像素

例如，对于4个标记的对象，列表将

{ (5,4), (5,5) , (6,4), (6,5) }

面积只是每个对象像素阵列的大小，但我如何才能找到周长，我是否应该再次迭代整个图像，查找单元的邻居，检查它是否是对象的角像素，或者有更简单的方法，仅基于坐标进行此操作

---

请建议最容易找到周长的方法，任何代码示例都将受到高度赞赏

尝试对您的图像进行广度优先搜索，（或者遍历您的点列表），并标记与不属于同一组的另一个像素相邻的每个像素

我不清楚您想要的周长是请求对象外部边缘上的每个像素，还是对象边界上的每个像素。我现在就假设前者

设置图像： 下面是你将如何着手做这件事。首先，将图像设置为二维阵列，每个像素都标有组号：

[ 0 0 0 0 1 1 0 ]
[ 0 0 2 0 0 0 0 ]
[ 0 2 2 0 3 3 0 ]
[ 0 2 2 0 0 3 0 ]
[ 0 2 2 0 0 0 0 ]
[ 0 2 0 0 4 4 0 ]
[ 0 0 0 0 4 4 0 ]

加载此文件的一个好方法是使用

扫描仪

对象逐个获取每个点：

List<Point> points = new ArrayList<>();
Scanner scanner = new Scanner( /* whatever your input source is */ );
String pointRegex = "\\(\\d,\\d\\)"; //looks for something like "(#,#)"
while(!scanner.hasNext(pointRegex)){
    String pointText = scanner.next(pointRegex); //For example, "(5,4)"
    Point point = getPointFromText(pointText); //turns a string into a point
    points.add(point);
}

这将对象（由我们的

点

对象集合“

点

”表示）放入一个

布尔

的网格中。我们只需要担心这一个对象，所以不需要加载任何其他对象。现在来看看周界上有哪些点

递归方法： 迭代法： 这与上面的非常相似，但直接跳到将周界点放入列表中

int[] deltaX = {-1,  0,  1, -1, 1, -1, 0, 1},
      deltaY = {-1, -1, -1,  0, 0,  1, 1, 1};
outer: for(Point p : points){
    inner: for(int i = 0 ; i < deltaX.length ; i++){
        int newX = p.getX() + deltaX[i];
        int newY = p.getY() + deltaY[i];
        //check if this new point we're considering is outside the image
        if(newX < 0 || newX >= width || newY<0 || newY>=height){
            perimeter.add(p); //if you decide bordering the edge of the image counts as being on the perimeter
            continue outer;
        }

        //check if this new point we're considering isn't part of the image
        if( binaryImage[p.getX()][p.getY()] != binaryImage[newX][newY] ){
            //if it isn't, then this Point p must be on the perimeter
            perimeter.add(p);
            continue outer;
        }
    }
}

int[]deltaX={-1,0,1，-1,1，-1,0,1}，
deltaY={1，-1，-1,0,0,1,1,1}；
外部：用于（点p：点）{
内部：for（int i=0；i=宽度| | newY=高度）{
add（p）；//如果您决定与图像的边缘相邻，则视为在周长上
继续对外开放；
}
//检查我们正在考虑的新点是否不是图像的一部分
if（binaryImage[p.getX（）][p.getY（）！=binaryImage[newX][newY]）{
//如果不是，那么点p必须在周界上
增加（p）；
继续对外开放；
}
}
}

注意标签

外部：

和

内部：

。这让我们可以选择在说

continue outer时跳过哪个for循环

好了！这将帮助您以二进制图像或列表的形式获取任何对象的周长

boolean[][] visitedBefore = new boolean[width][height];
boolean[][] isOnPerimeter = new boolean[width][height];
int[] deltaX = {-1,  0,  1, -1, 1, -1, 0, 1},
      deltaY = {-1, -1, -1,  0, 0,  1, 1, 1};
Queue<Point> searchNext = new LinkedList<>();
searchNext.add(points.get(0)); //Just need one point to get going
while(!searchNext.isEmpty()){
    Point p = searchNext.remove(); //take what's waiting at the front of the queue
    if(visitedBefore[p.getX()][p.getY()]){
        continue; //already check this spot!
    }

    //mark that we've been here
    visited[p.getX()][p.getY()] = true;

    //look at all of this Point's neighbors
    for(int i = 0 ; i < deltaX.length ; i++){
        int newX = p.getX() + deltaX[i];
        int newY = p.getY() + deltaY[i];

        //make sure this isn't out of bounds
        if(newX < 0 || newX >= width || newY<0 || newY>=height){
            isOnPerimeter[p.getX()][p.getY()] = true; //if you decide bordering the edge of the image counts as being on the perimeter
            continue;
        }

        //check if this new point we're considering isn't part of the image
        if( binaryImage[p.getX()][p.getY()] != binaryImage[newX][newY] ){
            //if it isn't, then this Point p must be on the perimeter
            isOnPerimeter[p.getX()][p.getY()] = true;
        } else {
            /* otherwise, this new point we're considering is part of the
             * same object, and could be part of the perimeter. */
            searchNext.add(new Point(newX, newY));
        }
    }
}

List<Point> perimeter = new ArrayList<Point>();
for(int x = 0 ; x < isOnPerimeter.length ; x++)
    for(int y = 0 ; y < isOnPerimeter[x].length ; y++)
        perimeter.add( new Point(x,y) );

int[] deltaX = {-1,  0,  1, -1, 1, -1, 0, 1},
      deltaY = {-1, -1, -1,  0, 0,  1, 1, 1};
outer: for(Point p : points){
    inner: for(int i = 0 ; i < deltaX.length ; i++){
        int newX = p.getX() + deltaX[i];
        int newY = p.getY() + deltaY[i];
        //check if this new point we're considering is outside the image
        if(newX < 0 || newX >= width || newY<0 || newY>=height){
            perimeter.add(p); //if you decide bordering the edge of the image counts as being on the perimeter
            continue outer;
        }

        //check if this new point we're considering isn't part of the image
        if( binaryImage[p.getX()][p.getY()] != binaryImage[newX][newY] ){
            //if it isn't, then this Point p must be on the perimeter
            perimeter.add(p);
            continue outer;
        }
    }
}