Three.js 3D缩放到鼠标位置

我有下一个问题。我有一个用Three.js+orbitControls.js制作的相机处理项目。有一个场景，上面有一个简单的立方体，还有相机。我有一台来自orbitControls.js的相机，当我使用鼠标滚轮时，它会朝屏幕中心放大/缩小（默认行为）。我需要修改它，这样就可以放大/缩小到3D对象上的鼠标位置，就像在3D编辑器中一样，比如3D Max、blender。

我确实理解逻辑应该是这样的：从开始缩放的位置，以及实际要放大的位置，存储该值，放大并获得新的鼠标位置，找到第一个鼠标位置和当前位置之间的偏移量，并在放大过程中将相机移动到该偏移量。但是，它工作不好。大多数情况下，我认为这可能是我的相机坐标的问题，以及我想放大的位置。

这是我的代码：

 window.onload = function() {
        var renderer = new THREE.WebGLRenderer();
        var width = 640;
        var height = 480;
        renderer.setSize( width, height );
        var div = document.getElementById('canvas')
        div.appendChild( renderer.domElement );
        var scene = new THREE.Scene();
            camera = new THREE.PerspectiveCamera(
            70,
            width / height,
            0.1,
            10000
        );
        camera.position.set( 0, 10, 10 );
        camera.lookAt( scene.position );
        control = new THREE.OrbitControls( camera );
        var geometry = new THREE.CubeGeometry( 5, 5, 5 );
        var material = new THREE.MeshLambertMaterial( { color: 0xFF0000 } );
         mesh = new THREE.Mesh( geometry, material );
        scene.add( mesh );
        var light = new THREE.PointLight( 0xFFFF00 );
        light.position.set( 10, 0, 10 );
        scene.add( light );
        var dt = 30;
        var renderFunc = function() {
          update(dt);
          renderer.render(scene,camera);
        };
        setInterval(renderFunc,dt);
        //MY ZOOMING CODE.
        var mouseVector = new THREE.Vector3();
        var projector = new THREE.Projector();  
        var startMousePos = new THREE.Vector3();
        var intersectedPoint = new THREE.Vector3();
        var intersectedOffset = new THREE.Vector3();
//FIND INTERSECTED POSITION ON THE CUBE         
        canvas.onmousemove = function (event)
        {
            mouseVector.x = ( event.clientX / 640 ) * 2 - 1;
            mouseVector.y = - ( event.clientY / 480 ) * 2 + 1;
            startMousePos.x = mouseVector.x;
            startMousePos.y = mouseVector.y;
            var vector = new THREE.Vector3( mouseVector.x, mouseVector.y, 1 );
            projector.unprojectVector( vector, camera );
            var ray = new THREE.Raycaster( camera.position, vector.sub( camera.position ).normalize() );
            var intersects = ray.intersectObjects( scene.children );
            if ( intersects.length > 0 )
            {
                 intersectedPoint = intersects[0].point;
            }
        }
//HERE I THINK I MUST TRANSLATE MY CAMERA.
        canvas.onmousewheel = function(event)
        {
            if(event.wheelDelta > 0)
            {
                camera.position.x -= intersectedPoint.x;
                camera.position.y -= intersectedPoint.y;
                camera.position.z -= intersectedPoint.z;
            }
            else
            {
                camera.position.x += intersectedPoint.x;
                camera.position.y += intersectedPoint.y;
                camera.position.z += intersectedPoint.z;
            }
        }
    };

以前有人这样做过吗？你能告诉我你是如何解决的吗？或者给我一个如何处理的建议？

谢谢！