Skip to content

GitLab

Switch branch/tag
Find file Normal viewHistoryPermalink
index.html 27.1 KB
Newer Older
allai5's avatar
add data and site info    
allai5 committed
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41 
<!DOCTYPE html> <html lang="en-US"> <head> <meta charset="UTF-8"> <meta http-equiv="X-UA-Compatible" content="IE=Edge"> <title>Simplification - </title> <link rel="shortcut icon" href="/favicon.ico" type="image/x-icon"> <link rel="stylesheet" href="/assets/css/just-the-docs-default.css"> <script type="text/javascript" src="/assets/js/vendor/lunr.min.js"></script> <script type="text/javascript" src="/assets/js/just-the-docs.js"></script> <meta name="viewport" content="width=device-width, initial-scale=1"> <!-- Begin Jekyll SEO tag v2.7.1 --> <title>Simplification</title> <meta name="generator" content="Jekyll v4.2.0" /> <meta property="og:title" content="Simplification" /> <meta property="og:locale" content="en_US" /> <meta name="twitter:card" content="summary" /> <meta property="twitter:title" content="Simplification" /> <script type="application/ld+json"> {"headline":"Simplification","@type":"WebPage","url":"/meshedit/global/simplify/","@context":"https://schema.org"}</script> <!-- End Jekyll SEO tag --> </head> <body> <svg xmlns="http://www.w3.org/2000/svg" style="display: none;"> <symbol id="svg-link" viewBox="0 0 24 24"> <title>Link</title> <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="feather feather-link"> <path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path> </svg> </symbol> <symbol id="svg-search" viewBox="0 0 24 24"> <title>Search</title> <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"> <circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line> </svg> </symbol> <symbol id="svg-menu" viewBox="0 0 24 24"> <title>Menu</title> <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="feather feather-menu"> <line x1="3" y1="12" x2="21" y2="12"></line><line x1="3" y1="6" x2="21" y2="6"></line><line x1="3" y1="18" x2="21" y2="18"></line> </svg> </symbol> <symbol id="svg-arrow-right" viewBox="0 0 24 24"> <title>Expand</title> <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="feather feather-chevron-right"> <polyline points="9 18 15 12 9 6"></polyline> </svg> </symbol> <symbol id="svg-doc" viewBox="0 0 24 24"> <title>Document</title> <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="feather feather-file"> <path d="M13 2H6a2 2 0 0 0-2 2v16a2 2 0 0 0 2 2h12a2 2 0 0 0 2-2V9z"></path><polyline points="13 2 13 9 20 9"></polyline> </svg> </symbol> </svg> <div class="side-bar"> <div class="site-header"> <a href="/" class="site-title lh-tight"> </a> <a href="#" id="menu-button" class="site-button"> <svg viewBox="0 0 24 24" class="icon"><use xlink:href="#svg-menu"></use></svg> </a> </div> <nav role="navigation" aria-label="Main" id="site-nav" class="site-nav"> <ul class="nav-list"><li class="nav-list-item"><a href="/" class="nav-list-link">Home</a></li><li class="nav-list-item"><a href="/git/" class="nav-list-link">GitHub Setup</a></li><li class="nav-list-item"><a href="/build/" class="nav-list-link">Building Scotty3D</a></li><li class="nav-list-item"><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/guide/" class="nav-list-link">User Guide</a><ul class="nav-list "><li class="nav-list-item "><a href="/guide/animate_mode/" class="nav-list-link">Animate</a></li><li class="nav-list-item "><a href="/guide/layout_mode/" class="nav-list-link">Layout</a></li><li class="nav-list-item "><a href="/guide/model_mode/" class="nav-list-link">Model</a></li><li class="nav-list-item "><a href="/guide/render_mode/" class="nav-list-link">Render</a></li><li class="nav-list-item "><a href="/guide/rigging_mode/" class="nav-list-link">Rig</a></li><li class="nav-list-item "><a href="/guide/simulate_mode/" class="nav-list-link">Simulate</a></li></ul></li><li class="nav-list-item active"><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/meshedit/" class="nav-list-link">A2: MeshEdit</a><ul class="nav-list "><li class="nav-list-item "><a href="/meshedit/halfedge" class="nav-list-link">Halfedge Mesh</a></li><li class="nav-list-item "><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/meshedit/local/" class="nav-list-link">Local Operations</a><ul class="nav-list"><li class="nav-list-item "> <a href="/meshedit/local/edge_flip" class="nav-list-link">Edge Flip Tutorial</a> </li><li class="nav-list-item "> <a href="/meshedit/local/bevel/" class="nav-list-link">Bevelling</a> </li></ul></li><li class="nav-list-item active"><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/meshedit/global/" class="nav-list-link">Global Operations</a><ul class="nav-list"><li class="nav-list-item "> <a href="/meshedit/global/catmull/" class="nav-list-link">Catmull-Clark Subdivision</a> </li><li class="nav-list-item "> <a href="/meshedit/global/remesh/" class="nav-list-link">Isotropic Remeshing</a> </li><li class="nav-list-item "> <a href="/meshedit/global/linear/" class="nav-list-link">Linear Subdivision</a> </li><li class="nav-list-item "> <a href="/meshedit/global/loop/" class="nav-list-link">Loop Subdivision</a> </li><li class="nav-list-item active"> <a href="/meshedit/global/simplify/" class="nav-list-link active">Simplification</a> </li><li class="nav-list-item "> <a href="/meshedit/global/triangulate/" class="nav-list-link">Triangulation</a> </li></ul></li></ul></li><li class="nav-list-item"><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/pathtracer/" class="nav-list-link">A3: Pathtracer</a><ul class="nav-list "><li class="nav-list-item "><a href="/pathtracer/camera_rays" class="nav-list-link">(Task 1) Camera Rays</a></li><li class="nav-list-item "><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/pathtracer/intersecting_objects" class="nav-list-link">(Task 2) Intersections</a><ul class="nav-list"><li class="nav-list-item "> <a href="/pathtracer/ray_triangle_intersection" class="nav-list-link">Ray Triangle Intersection</a> </li><li class="nav-list-item "> <a href="/pathtracer/ray_sphere_intersection" class="nav-list-link">Ray Sphere Intersection</a> </li></ul></li><li class="nav-list-item "><a href="/pathtracer/bounding_volume_hierarchy" class="nav-list-link">(Task 3) BVH</a></li><li class="nav-list-item "><a href="/pathtracer/shadow_rays" class="nav-list-link">(Task 4) Shadow Rays</a></li><li class="nav-list-item "><a href="/pathtracer/path_tracing" class="nav-list-link">(Task 5) Path Tracing</a></li><li class="nav-list-item "><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/pathtracer/materials" class="nav-list-link">(Task 6) Materials</a><ul class="nav-list"><li class="nav-list-item "> <a href="/pathtracer/dielectrics_and_transmission" class="nav-list-link">Dielectrics and Transmission</a> </li></ul></li><li class="nav-list-item "><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/pathtracer/environment_lighting" class="nav-list-link">(Task 7) Environment Lighting</a><ul class="nav-list"><li class="nav-list-item "> <a href="/pathtracer/importance_sampling" class="nav-list-link">Environment Light Importance Sampling</a> </li></ul></li><li class="nav-list-item "><a href="/pathtracer/visualization_of_normals" class="nav-list-link">Visualization of normals</a></li></ul></li><li class="nav-list-item"><a href="#" class="nav-list-expander"><svg viewBox="0 0 24 24"><use xlink:href="#svg-arrow-right"></use></svg></a><a href="/animation/" class="nav-list-link">A4: Animation</a><ul class="nav-list "><li class="nav-list-item "><a href="/animation/splines" class="nav-list-link">Splines</a></li><li class="nav-list-item "><a href="/animation/skeleton_kinematics" class="nav-list-link">Skeleton Kinematics</a></li><li class="nav-list-item "><a href="/animation/skinning" class="nav-list-link">Skinning</a></li><li class="nav-list-item "><a href="/animation/particles" class="nav-list-link">Particles</a></li></ul></li></ul> </nav> <footer class="site-footer"> This site uses <a href="https://github.com/pmarsceill/just-the-docs">Just the Docs</a>, a documentation theme for Jekyll. </footer> </div> <div class="main" id="top"> <div id="main-header" class="main-header"> <div class="search"> <div class="search-input-wrap"> <input type="text" id="search-input" class="search-input" tabindex="0" placeholder="Search " aria-label="Search " autocomplete="off"> <label for="search-input" class="search-label"><svg viewBox="0 0 24 24" class="search-icon"><use xlink:href="#svg-search"></use></svg></label> </div> <div id="search-results" class="search-results"></div> </div> </div> <div id="main-content-wrap" class="main-content-wrap"> <nav aria-label="Breadcrumb" class="breadcrumb-nav"> <ol class="breadcrumb-nav-list"> <li class="breadcrumb-nav-list-item"><a href="/meshedit/">A2: MeshEdit</a></li> <li class="breadcrumb-nav-list-item"><a href="/meshedit/global/">Global Operations</a></li> <li class="breadcrumb-nav-list-item"><span>Simplification</span></li> </ol> </nav> <div id="main-content" class="main-content" role="main"> <h1 id="simplification"> <a href="#simplification" class="anchor-heading" aria-labelledby="simplification"><svg viewBox="0 0 16 16" aria-hidden="true"><use xlink:href="#svg-link"></use></svg></a> Simplification </h1> <p><img src="quad_simplify.png" alt="Surface simplification via quadric error metric" /></p> <p>For an in-practice example, see the <a href="/Scotty3D/guide/model">User Guide</a>.</p> <p>Just as with images, meshes often have far more samples than we really need. The simplification method in Scotty3D simplifies a given triangle mesh by applying <em>quadric error simplification</em> (note that this method is for <strong>triangle meshes only</strong>!). This method was originally developed at CMU by Michael Garland and Paul Heckbert, in their paper <a href="http://www.cs.cmu.edu/~./garland/quadrics/quadrics.html">Surface Simplification Using Quadric Error Metrics</a>. (Looking at this paper – or the many slides and presentations online that reference it – may be very helpful in understanding and implementing this part of the assignment!)</p> <p>The basic idea is to iteratively collapse edges until we reach the desired number of triangles. The more edges we collapse, the simpler the mesh becomes. The only question is: which edges should we collapse? And where should we put the new vertex when we collapse an edge? Finding the sequence of edge collapses (and vertex positions) that give an <em>optimal</em> approximation of the surface would be very difficult – likely impossible! Garland and Heckbert instead proposed a simple, greedy scheme that works quite well in practice, and is the basis of many mesh simplification tools today. Roughly speaking, we’re going to write down a function that measures the distance to a given triangle, and then “accumulate” this function as many triangles get merged together.</p> <p>More precisely, we can write the distance d of a point <em>x</em> to a plane with normal <em>N</em> passing through a point <em>p</em> as dist(<em>x</em>) = dot(<em>N</em>, <em>x</em> - <em>p</em>)</p> <center><img src="plane_normal.png" style="height:360px" /></center> <p>In other words, we measure the extent of the vector from <em>p</em> to <em>x</em> along the normal direction. This quantity gives us a value that is either <em>positive</em> (above the plane), or <em>negative</em> (below the plane). Suppose that <em>x</em> has coordinates (<em>x</em>,<em>y</em>,<em>z</em>), <em>N</em> has coordinates (<em>a</em>,<em>b</em>,<em>c</em>), and let <em>d</em>(<em>x</em>) = -dot(<em>N</em>, <em>p</em>), then in <em>homogeneous</em> coordinates, the distance to the plane is just</p> <p>dot(<em>u</em>, <em>v</em>)</p> <p>where <em>u</em> = (<em>x</em>,<em>y</em>,<em>z</em>,<em>1</em>) and <em>v</em> = (<em>a</em>,<em>b</em>,<em>c</em>,<em>d</em>). When we’re measuring the quality of an approximation, we don’t care whether we’re above or below the surface; just how <em>far away</em> we are from the original surface. Therefore, we’re going to consider the <em>square</em> of the distance, which we can write in homogeneous coordinates as</p> <center><img src="homogeneous_coord.png" style="height:40px" /></center> <p>where T denotes the transpose of a vector. The term <em>vv</em>^T is an <a href="https://en.wikipedia.org/wiki/Outer_product">outer product</a> of the vector <em>v</em> with itself, which gives us a symmetric matrix <em>K</em> = <em>vv</em>^T. In components, this matrix would look like</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>a^2   ab    ac   ad
ab    b^2   bc   bd
ac    bc    c^2  cd
ad    bd    cd   d^2
</code></pre></div></div> <p>but in Scotty3D it can be constructed by simply calling the method <code class="language-plaintext highlighter-rouge">outer( Vec4, Vec4 )</code> in <code class="language-plaintext highlighter-rouge">lib/mat4.h</code> that takes a pair of vectors in homogeneous coordinates and returns the outer product as a 4x4 matrix. We will refer to this matrix as a “quadric,” because it also describes a <a href="https://en.wikipedia.org/wiki/Quadric">quadric surface</a>.</p> <p>The matrix <em>K</em> tells us something about the distance to a plane. We can also get some idea of how far we are from a <em>vertex</em> by considering the sum of the squared distances to the planes passing through all triangles that touch that vertex. In other words, we will say that the distance to a small neighborhood around the vertex i can be approximated by the sum of the quadrics on the incident faces ijk:</p> <center><img src="K_sum.png" style="height:100px" /></center> <center><img src="vert_normals.png" style="height:360px" /></center> <p>Likewise, the distance to an <em>edge</em> ij will be approximated by the sum of the quadrics at its two endpoints:</p> <center><img src="edge_k_sum.png" style="height:50px" /></center> <p>The sums above should then be easy to compute – you can just add up the <code class="language-plaintext highlighter-rouge">Mat4</code> objects around a vertex or along an edge using the usual “+” operator. You do not need to write an explicit loop over the 16 entries of the matrix.</p> <p>Once you have a quadric <em>K</em> associated with an edge <em>ij</em>, you can ask the following question: if we collapse the edge to a point <em>x</em>, where should we put the new point so that it minimizes the (approximate) distance to the original surface? In other words, where should it go so that it minimizes the quantity <em>x</em>^T <em>K x</em>?</p> <p>Just like any other function, we can look for the minimum by taking the derivative with respect to <em>x</em> and setting it equal to zero. (By the way, in this case we’re always going to get a <em>minimum</em> and not a <em>maximum</em> because the matrices K are all <a href="https://en.wikipedia.org/wiki/Positive-definite_matrix">positive-definite</a>.) In other words, we want to solve the small (4x4) linear system</p> <p><em>K u</em> = <em>0</em></p> <p>for the optimal position <em>u</em>, expressed in homogeneous coordinates. We can simplify this situation a bit by remembering that the homogeneous coordinate for a point in 3-space is just 1. After a few simple manipulations, then, we can rewrite this same system as an even smaller 3x3 linear system</p> <p><em>Ax</em> = <em>b</em></p> <p>where A is the upper-left 3x3 block of K, and b is <em>minus</em> the upper-right 3x1 column. In other words, the entries of A are just</p> <center><img src="K_A_block.png" style="height:100px" /></center> <p>and the entries of b are</p> <center><img src="b_vec.png" style="height:100px" /></center> <p>The cost associated with this solution can be found by plugging <em>x</em> back into our original expression, i.e., the cost is just</p> <p><em>x</em>^T <em>K</em> <em>x</em></p> <p>where <em>K</em> is the quadric associated with the edge. Fortunately, <em>you do not need to write any code to solve this linear system</em>. It can be solved using the method <code class="language-plaintext highlighter-rouge">Mat4::inverse()</code> which computes the inverse of a 4x4 matrix. Note that while we really want to work with a 3x3 matrix here, using the upper left 3x3 block of a 4x4 matrix is equivalent, given that the 4th row/column remain as in the identity matrix. In particular, you can write something like this:</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>Mat4 A; // computed by accumulating quadrics and then extacting the upper-left 3x3 block
Vec3 b;  // computed by extracting minus the upper-right 3x1 column from the same matrix
Vec3 x = A.inverse() * b; // solve Ax = b for x by hitting both sides with the inverse of A
</code></pre></div></div> <p>However, A might not always be invertible: consider the case where the mesh is composed of points all on the same plane. In this case, you need to select an optimal point along the original edge. Please read <a href="http://reports-archive.adm.cs.cmu.edu/anon/1999/CMU-CS-99-105.pdf">Garland’s paper</a> on page 62 section 3.5 for more details.</p> <p>If you’re a bit lost at this point, don’t worry! There are a lot of details to go through, and we’ll summarize everything again in the implementation section. The main idea to keep in mind right now is:</p> <ul> <li>we’re storing a matrix at every vertex that encodes (roughly) the distance to the surface, and</li> <li>for each edge, we want to find the point that is (roughly) as close as possible to the surface, according to the matrices at its endpoints.</li> </ul> <p>As we collapse edges, the matrices at endpoints will be combined by just adding them together. So, as we perform more and more edge collapses, these matrices will try to capture the distance to a larger and larger region of the original surface.</p> <p>The one final thing we want to think about is performance. At each iteration, we want to collapse the edge that results in the <em>least</em> deviation from our original surface. But testing every edge, every single iteration sounds pretty expensive! (Something like O(n^2).) Instead, we’re going to put all our edges into a <a href="https://en.wikipedia.org/wiki/Priority_queue">priority queue</a> that efficiently keeps track of the “best” edge for us, even as we add and remove edges from our mesh. In the code framework, we actually introduce a new class called an <code class="language-plaintext highlighter-rouge">Edge_Record</code> that encodes all the essential information about our edge:</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>// An edge record keeps track of all the information about edges
// that we need while applying our mesh simplification algorithm.
class Edge_Record {
public:
    Edge_Record() {}
    Edge_Record(std::unordered_map&lt;Halfedge_Mesh::VertexRef, Mat4&gt;&amp; vertex_quadrics,
                Halfedge_Mesh::EdgeRef e) : edge(e) {

        // The second constructor takes a dictionary mapping vertices
        // to quadric error matrices and an edge reference. It then
        // computes the sum of the quadrics at the two endpoints
        // and solves for the optimal midpoint position as measured
        // by this quadric. It also stores the value of this quadric
        // as the "score" used by the priority queue.
    }

    EdgeRef edge; // the edge referred to by this record

    Vec3 optimal; // the optimal point, if we were
                  // to collapse this edge next

    float cost; // the cost associated with collapsing this edge,
                // which is very (very!) roughly something like
                // the distance we'll deviate from the original
                // surface if this edge is collapsed
};
</code></pre></div></div> <p>Within <code class="language-plaintext highlighter-rouge">Halfedge_Mesh::simplify</code>, you will create a dictionary <code class="language-plaintext highlighter-rouge">vertex_quadrics</code> mapping vertices to quadric error matrices. We will use a <code class="language-plaintext highlighter-rouge">std::unordered_map</code> for this purpose, which is the hash map provided by the STL. Its usage is detailed in the <a href="https://en.cppreference.com/w/cpp/container/unordered_map">C++ documentation</a>. To initialize the record for a given edge <code class="language-plaintext highlighter-rouge">e</code>, you can use this dictionary to write</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>Edge_Record record(vertex_quadrics, e);
</code></pre></div></div> <p>Similarly to how we created a dictionary mapping vertices to quadric matrices, we will also want to associate this record with its edge using the <code class="language-plaintext highlighter-rouge">edge_records</code> dictionary:</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>edge_records[e] = record;
</code></pre></div></div> <p>Further, we will want to add the record to a priority queue, which is always sorted according to the cost of collapsing each edge. The starter code also provides the helper class <code class="language-plaintext highlighter-rouge">PQueue</code> for this purpose. For example:</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>PQueue&lt;Edge_Record&gt; queue;
queue.insert(record);
</code></pre></div></div> <p>If we ever want to know what the best edge is to collapse, we can just look at the top of the priority queue:</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>Edge_Record bestEdge = queue.top();
</code></pre></div></div> <p>More documentation is provided inline in <code class="language-plaintext highlighter-rouge">student/meshedit.cpp</code>.</p> <p>Though conceptually sophisticated, quadric error simplification is actually not too hard to implement. It basically boils down to two methods:</p> <div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>Edge_Record::Edge_Record(std::unordered_map&lt;Halfedge_Mesh::VertexRef, Mat4&gt;&amp; vertex_quadrics, EdgeIter e);
Halfedge_Mesh::simplify();
</code></pre></div></div> <p>As discussed above, the edge record initializer should:</p> <ol> <li>Compute a quadric for the edge as the sum of the quadrics at endpoints.</li> <li>Build a 3x3 linear system for the optimal collapsed point, as described above.</li> <li>Solve this system and store the optimal point in <code class="language-plaintext highlighter-rouge">Edge_Record::optimal</code>.</li> <li>Compute the corresponding error value and store it in <code class="language-plaintext highlighter-rouge">Edge_Record::cost</code>.</li> <li>Store the edge in <code class="language-plaintext highlighter-rouge">Edge_Record::edge</code>.</li> </ol> <p>The downsampling routine can then be implemented by following this basic recipe:</p> <ol> <li>Compute quadrics for each face by simply writing the plane equation for that face in homogeneous coordinates, and building the corresponding quadric matrix using <code class="language-plaintext highlighter-rouge">outer()</code>. This matrix should be stored in the yet-unmentioned dictionary <code class="language-plaintext highlighter-rouge">face_quadrics</code>.</li> <li>Compute an initial quadric for each vertex by adding up the quadrics at all the faces touching that vertex. This matrix should be stored in <code class="language-plaintext highlighter-rouge">vertex_quadrics</code>. (Note that these quadrics must be updated as edges are collapsed.)</li> <li>For each edge, create an <code class="language-plaintext highlighter-rouge">Edge_Record</code>, insert it into the <code class="language-plaintext highlighter-rouge">edge_records</code> dictionary, and add it to one global <code class="language-plaintext highlighter-rouge">PQueue&lt;Edge_Record&gt;</code> queue.</li> <li>Until a target number of triangles is reached, collapse the best/cheapest edge (as determined by the priority queue) and set the quadric at the new vertex to the sum of the quadrics at the endpoints of the original edge. You will also have to update the cost of any edge connected to this vertex.</li> </ol> <p>The algorithm should terminate when a target number of triangles is reached – for the purpose of this assignment, you should set this number to 1/4th the number of triangles in the input (since subdivision will give you a factor of 4 in the opposite direction). Note that to <em>get</em> the best element from the queue you call <code class="language-plaintext highlighter-rouge">PQueue::top()</code>, whereas to <em>remove</em> the best element from the top you must call <code class="language-plaintext highlighter-rouge">PQueue::pop()</code> (the separation of these two tasks is fairly standard in STL-like data structures).</p> <p>As with subdivision, it is critical that you carefully reason about which mesh elements get added/deleted in what order – particularly in Step 4. A good way to implement Step 4 would be:</p> <ol> <li>Get the cheapest edge from the queue.</li> <li><strong>Remove the cheapest edge from the queue by calling <code class="language-plaintext highlighter-rouge">pop()</code>.</strong></li> <li>Compute the new quadric by summing the quadrics at its two endpoints.</li> <li><strong>Remove any edge touching either of its endpoints from the queue.</strong></li> <li>Collapse the edge.</li> <li>Set the quadric of the new vertex to the quadric computed in Step 3.</li> <li><strong>Insert any edge touching the new vertex into the queue, creating new edge records for each of them.</strong></li> </ol> <p>Steps 4 and 7 are highlighted because it is easy to get these steps wrong. For instance, if you collapse the edge first, you may no longer be able to access the edges that need to be removed from the queue.</p> <p>A working implementation should look something like the examples below. You may find it easiest to implement this algorithm in stages. For instance, <em>first</em> get the edge collapses working, using just the edge midpoint rather than the optimal point, <em>then</em> worry about solving for the point that minimizes quadric error.</p> <!--![Quadric error simplification examples](quad_example.png)--> <center><img src="quad_example.png" style="height:480px" /></center> </div> </div> <div class="search-overlay"></div> </div> </body> </html>