经纬度演变的轮廓缩放问题？

  • Z时代
  • 2024-03-11
  • 分类：IT

目标是实现上方的算法（网上的图片）

按照下面的方式进行计算 但并没有达到我的预期（感觉是经纬度转墨卡托出了问题）

 public List<com.vividsolutions.jts.geom.Coordinate> warpAndWeft() {
        //拿到了经纬度
//
        List<com.vividsolutions.jts.geom.Coordinate> list = new ArrayList<>();
        com.vividsolutions.jts.geom.Coordinate coordinate = new com.vividsolutions.jts.geom.Coordinate(120.0595083689628, 30.313006982475443);
        com.vividsolutions.jts.geom.Coordinate coordinate2 = new com.vividsolutions.jts.geom.Coordinate(120.05956464781273, 30.312856394593823);
        com.vividsolutions.jts.geom.Coordinate coordinate3 = new com.vividsolutions.jts.geom.Coordinate(120.05976961479806, 30.31289540653132);
        com.vividsolutions.jts.geom.Coordinate coordinate4 = new com.vividsolutions.jts.geom.Coordinate(120.05972807069857, 30.313059860492974);
        com.vividsolutions.jts.geom.Coordinate coordinate5 = new com.vividsolutions.jts.geom.Coordinate(120.05964635634035, 30.31299054114971);
        com.vividsolutions.jts.geom.Coordinate coordinate6 = new com.vividsolutions.jts.geom.Coordinate(120.0595083689628, 30.313006982475443);
        list.add(coordinate);
        list.add(coordinate2);
        list.add(coordinate3);
        list.add(coordinate4);
        list.add(coordinate5);
        list.add(coordinate6);
        int size = list.size();
        //记录结果
        List<com.vividsolutions.jts.geom.Coordinate> result = new ArrayList<>();
        for (int i = 0; i < size; i++) {
            GeoTransform geoTransform = new GeoTransform();
            //将经纬度转变成二维坐标下(采用同高德一致 墨卡托)
            com.vividsolutions.jts.geom.Coordinate coordinates = geoTransform.geographic2Mercator(list.get(i));
            Point point = new Point();
            point.setLongitude(coordinates.x);
            point.setLatitude(coordinates.y);
            com.vividsolutions.jts.geom.Coordinate coordinates1 = geoTransform.geographic2Mercator(list.get(i == 0 ? size - 1 : i - 1));
            Point point1 = new Point();
            point1.setLongitude(coordinates1.x);
            point1.setLatitude(coordinates1.y);
            com.vividsolutions.jts.geom.Coordinate coordinates2 = geoTransform.geographic2Mercator(list.get(i == size - 1 ? 0 : i + 1));
            Point point2 = new Point();
            point2.setLongitude(coordinates2.x);
            point2.setLatitude(coordinates2.y);
            //向量pp1
            Point sub = VectorUtil.sub(point1, point);//pp1
            System.out.println(point1.getLongitude()-point.getLongitude());
            double norm = VectorUtil.norm(sub); //  |pp1|  膜
            double normX1 = sub.getLongitude() / norm;  //x向量单位化  v
            double normY1 = sub.getLatitude() / norm;  //y向量单位化   v
            Point pointV3 = new Point(normX1, normY1);
            //向量pp2
            Point sub2 = VectorUtil.sub(point2, point);//pp2
            double norm2 = VectorUtil.norm(sub2); //  |pp2|  膜
            double normX2 = sub2.getLongitude() / norm2;  //x向量单位化  v2
            double normY2 = sub2.getLatitude() / norm2;  //y向量单位化   v2
            Point pointV4 = new Point(normX2, normY2);
            //pQ的距离 |pQ| =L / Math.sqrt(1 - (v1x * v2x + v1y * v2y) / 2)
            double vectorLen =6 / Math.sqrt((1 - (VectorUtil.cross(pointV3, pointV4))) / 2);
            // 根据向量的叉乘积来判断角是凹角还是凸角 (sub.getLongitude() * sub2.getLatitude()) + (-1 * sub.getLatitude() * sub2.getLongitude())
            if ((VectorUtil.cross(sub,sub2)) < 0) {
                pointV4.setLongitude(-1 * pointV4.getLongitude());
                pointV4.setLatitude(-1 * pointV4.getLatitude());
                pointV3.setLongitude(-1 * pointV3.getLongitude());
                pointV3.setLatitude(-1 * pointV3.getLatitude());
            }
            //pQ的方向
        /*    double vectorX = pointV3.getLongitude() + pointV4.getLongitude(); //  PQ的向量=v1+v2(向量加  四边形法则)
            double vectorY = pointV3.getLatitude() + pointV4.getLatitude(); //  PQ的向量=v1+v2(向量加  四边形法则)
            Point point3 = new Point(vectorX, vectorY);*/
            Point point3 = VectorUtil.add(pointV3, pointV4);
            double n = vectorLen / VectorUtil.norm(point3);
            double vectorUnitX = n * point3.getLongitude();
            double vectorUnitY = n *  point3.getLatitude();
            //com.vividsolutions.jts.geom.Coordinate coordinate1 = new com.vividsolutions.jts.geom.Coordinate(vectorUnitX, vectorUnitY);
com.vividsolutions.jts.geom.Coordinate coordinate1 = new com.vividsolutions.jts.geom.Coordinate(vectorUnitX+ point.geteLongitude();
            , vectorUnitY+point.getLatitude());
            System.out.println(coordinate1.toString());
//这里是问题出现的地方，得到计算后的墨卡托坐标后转不回了经纬度 至于为什么转不回还没搞清楚
            com.vividsolutions.jts.geom.Coordinate theWarpAndWeftAfterChange = geoTransform.mercator2Geographic(coordinate1);
            result.add(theWarpAndWeftAfterChange);
        }
        return result;
    }

结果: 这是不是说明我转墨卡托计算 然后数智超过了计算机的位数（大概这意思：因为我看到了NaN） 那推论 计算后的数值太大所以墨卡托的点转变不会经纬度 这样对吗

[[NaN, NaN], [-2.685147104214853E-5, -4.6804572288379725E-5], [4.539105157497984E-5, -2.9064574458363977E-5], [1.6026060195087032E-5, 5.37608295495793E-5], [-5.2514021949098855E-5, 1.585992833204243E-4], [NaN, NaN]]

回答：

可能是因为你起始点相同，导致起始点的计算涉及到了0的除法，出现了NaN

以上是 经纬度演变的轮廓缩放问题？ 的全部内容， 来源链接： utcz.com/p/945059.html

回到顶部