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
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
|
<script src="./polygon_polygon/dist/output.js"></script>
<script>
window.onload = function() {
var lPlayElement = document.getElementById('gl_canvas_play'),
lStopElement = document.getElementById('gl_canvas_stop');
lPlayElement.addEventListener('click', function() {
lPlayElement.style.display = 'none';
lStopElement.style.display = 'block';
});
lStopElement.addEventListener('click', function() {
lStopElement.style.display = 'none';
lPlayElement.style.display = 'block';
});
}
</script>
<article>
<h1>Separating Axis Theorem</h1>
<section>
<p>
The Separating Axis Theorem (SAT) provides a way to find the intersection between any <i>n</i>-sided <a href='https://ianqvist.blogspot.com/2009/09/convex-polygon-based-collision.html'>convex</a> polygon or circle. In this tutorial, I will explain how this theorem works, and how you can use it to both detect and resolve collisions in your simulation.
</p>
<ul>
<li><a href='#explanation_of_separating_axis_theorem'>Explanation of Separating Axis Theorem</a></li>
<li><a href='#algorithm_for_finding_intersection'>Algorithm for Finding the Intersection</a></li>
<li><a href='#collision_resolution'>SAT Collision Resolution</a></li>
<ul>
<li><a href='#intersecting_edge'>Finding the Intersecting Edge</a></li>
<li><a href='#intersecting_point'>Finding the Intersecting Point</a></li>
</ul>
<li><a href='#algorithm_for_collision_resolution'>Algorithm for Collision Resolution</a></li>
<li><a href='#live_example'>Live Example</a></li>
</ul>
</section>
<section>
<h2 id="#explanation_of_separating_axis_theorem">Explanation of Separating Axis Theorem</h2>
<p>
SAT makes use of vector projection to figure out whether or not two concave polygons are intersecting. I believe the best way to describe it is to show you the instance where two polygons do <b>not</b> intersect.
<br/>
<br/>
Images two shapes <b>A</b> and <b>B</b> as shown below (they are triangles here, but they can be anything really).
<br/>
<br />
<div class='image'>
<img src="polygon_polygon/images/1.png" />
</div>
<br/>
We know inutitively that if we were able to draw a line between the two objects, that means that they're not intersecting. In this case, that line is a vertical-ish line:
<br/>
<br />
<div class='image'>
<img src='polygon_polygon/images/2.png' />
</div>
<br/>
That line could be any of the infinite number of lines between the objects, so we can't really select that <i>exact</i> line. So the problem becomes: how do we <i>infer</i> that the vertical line shown in the previous diagram is a line that splits the objects? That is where the Separating Axis Theorem comes in.
<br />
<br/>
To do separating axis theorem, we iterate all of the edges of <b>both</b> shapes. For each edge, we get the normal at that edge (In two-dimensions, this amounts to getting the perpendicular of the line segment defined by the two points of the edge).
<br />
<br />
<div class='image'>
<img src='polygon_polygon/images/2a.png' />
</div>
<br />
We then, in our heads, extend this normal to create an axis across our scene (keep in mind that you won't have to do any "extending" of this axis in code; the normal will do just fine).
<div class='image'>
<img src='polygon_polygon/images/2d.png' />
</div>
<br />
At this point, you will see that I named the three vertices on each of the triangles. Triangle <b>A</b> is made of vertices <i>a1, a2, a3</i> and triangle <b>B</b> is made of vertices <i>b1, b2, b3</i>. At this point, we project the shape onto the axis. To do this, we project the vertices of each shape onto the axis to figure out which one <i>maximizes</i> the dot product and which one <i>minimizes</i> the dot product. The projection of the shape is then the range defined by this min and max value.
<br />
<br />
<div class='image'>
<img src='polygon_polygon/images/2e.png' />
</div>
<br />
Finally, we check if the two ranges overlap. If they don't overlap, then the shapes do <b>not</b> intersect. If they do overlap, we repeat the process for all of the other edges. If all of these ranges overlap, then the shape overlaps, and there is no separating line between them.
<br />
<br />
<div class='image'>
<img src='polygon_polygon/images/2f.png' />
</div>
<br />
</p>
</section>
<section>
<h2 id='algorithm_for_finding_intersection'>Algorithm for Finding the Intersection</h2>
<p>
Given two polygons <b>A</b> and <b>B</b>:
<ol>
<li>For each edge on <b>A</b>, get the normal <i>n</i> of that edge.</li>
<li>Project each vertex <i>v</i> of <b>A</b> onto <i>n</i>. Return the minimum and maximum projection of all vertices as the projection of the shape.</li>
<li>Repeat Step 2 for polygon <b>B</b>.</li>
<li>If the min and max projections found in Steps 2 and 3 do <b>NOT</b> overlap, the polygons are not intersecting. Return false.</li>
<li>If the projections overlap for each edge of both shapes, the shapes are intersecting. Return true.</li>
</ol>
And that is all there is to <i>finding</i> the intersection between two convex polygons. Here is the code snippet that does this if you are interested:
#SNIPPET polygon_polygon/snippet1.cpp
</p>
</section>
<section>
<h2 id='collision_resolution'>SAT Collision Resolution</h2>
<p>
Now that we know that our objects have intersected, we want to send them tumbling away from one another in order to simulate a collision. To do this, we will need to find the following things:
<ul>
<li><b>Point of Application</b>: at what point on each object did the objects first intersect</li>
<li><b>Collision Normal</b>: in what direction, point towards object <b>A</b>, did the polygons intersect</li>
<li><b>Relative Velocity</b>: easily found by taking the difference between the two velocities.
</ul>
<h3 id='intersecting_edge'>Finding the Intersecting Edge</h3>
We can already figure out that the following two triangles intersect one another:
<br />
<br />
<div class='image'>
<img src='polygon_polygon/images/3a.png' />
</div>
<br/><br/>
We know that <b>A</b> can only intersect <b>B</b> if: (1) a vertex from <b>A</b> is inside of <b>B</b>, (2) an edge of <b>A</b> flatly intersects an edge of <b>B</b>, or (3) a vertex of <b>A</b> overlaps exactly a vertex of <b>B</b>. For our purposes, scenario 3 is quite unlikely, so we can ignore it for now. We will start with the first case, since it is the most likely.
<br /> <br />
The following drawing shows both of our objects projected onto the axis defined by object <b>B</b>'s leftmost edge:
<br /> <br />
<div class='image'>
<img src='polygon_polygon/images/3c.png' />
</div>
<br/><br/>
If you were to repeat this exercise of projecting the shapes onto every axis of both shapes, you will begin to unconver an interesting truth: <i>The intersecting edge will be the one where the projection of triangle <b>A</b> overlaps with the projection of triangle <b>B</b> the <b>least</b></i>! This should be unsurprising if you consider it. If our simulation is running quickly enough, the point of object <b>A</b> that is penetrating the edge of object <b>B</b> will just <i>barely</i> be inside of <b>B</b>. At the same time, it <i>should</i> be the only point that is currently inside of B. Hence, the overlap of the projection on that axis will be from that edge of <b>B</b> to that point on <b>A</b>. Knowing that this edge caused the intersection, we now have the collision normal.
<br />
<h3 id='intersecting_point'>Finding which point causes the intersection</h3>
This part is made easy by the fact that we know which edge of object <b>B</b> caused the intersection in the first place. We know, then, that the intersecting point on object <b>A</b> must be the point that is currently inside of <b>B</b>. Simply put, this point will be the vertex of <b>A</b> that was projected onto that axis that is <i>closest</i> to the edge of <b>B</b>. This will be true so long as the simulation is properly fast and the object is not infinitesmley small (in which case, we're dealing with particle physics, and you shouldn't be using this tutorial).
</p>
</section>
<section>
<h2 id='algorithm_for_collision_resolution'>Algorithm for Collision Resolution</h2>
<p>
Given two polygons <b>A</b> and <b>B</b>:
<ol>
<li>Find the intersecton as before</li>
<li>While finding the intersection, keep track of the axis that the projection on object <b>A</b> overlapped with the corresponding projection on object <b>B</b> the least. Remember the edge that this that held this axis.</li>
<li>Using this axis, find out which point on the other object inside of the edge by comparing the distances from the two points on the edge to the two points projected onto the axis created by that edge.</li>
<li>Use the calcualted edge to find the normal and the overlap point to find the point of application.</li>
</ol>
And that is it! Here is the code for this algorithm:
#SNIPPET polygon_polygon/snippet2.cpp
</p>
</section>
<section>
<h2 id='live_example'>
Live Example of Intersection Detection
</h2>
<p>
A live example of Separating Axis Theorem running with full collision detection. Click <button id='reverse_gravity'>this button</button>to reverse gravity on the simualtion.
</p>
<div class="opengl_canvas_container">
<canvas id="gl_canvas" width="800" height="600"></canvas>
<button id="gl_canvas_play" class="play_button">
Play
</button>
<button id="gl_canvas_stop" class="stop_button">
Stop
</button>
</div>
</section>
<footer id="references">
<h2>References</h2>
<ul>
<li><a href="https://en.wikipedia.org/wiki/Vector_projection">Vector Projection Wikapedia</a></li>
</ul>
</footer>
</article>
|
