I have tried to create NPC character that can "see" the player by using cones of vision.
The NPC will rotate back and forth at all times.
My problem is that the arc has a generic and unchanging position, but when its drawn to the screen it looks correct.
[Screenshots of the collisions in action][1]
[GitHub link for java files][2]
I'm using Arc2D to draw the shape like this in my NPC class
// Update the shapes used in the npc
rect.setRect(x, y, w, h);
ellipse.setFrame(rect);
visionArc.setArcByCenter(cx, cy, visionDistance, visionAngle, visionAngle * 2, Arc2D.PIE);
/ CenterX, CenterY (of the npc),
/ the distance from the arc to the npc
/ a constant value around 45 degrees and a constant value around 90 degress (to make a pie shape)
I've tried multiplying the position and the angles by the sin and cosine of the NPC's current angle
something like these
visionArc.setArcByCenter(cx * (Math.cos(Math.toRadians(angle))), cy (Math.sin(Math.toRadians(angle)), visionDistance, visionAngle, visionAngle * 2, Arc2D.PIE);
or
visionArc.setArcByCenter(cx, cy, visionDistance, visionAngle - angle, (visionAngle + angle) * 2, Arc2D.PIE);
or
visionArc.setArcByCenter(cx, cy, visionDistance, visionAngle * (Math.cos(Math.toRadians(angle))), visionAngle * 2, Arc2D.PIE);
I've tried a lot but can't seem to find what works. Making the vision angles not constant makes an arc that expands and contracts, and multiplying the position by the sin or cosine of the angle will make the arc fly around the screen, which doesn't really work either.
This is the function that draws the given NPC
public void drawNPC(NPC npc, Graphics2D g2, AffineTransform old) {
// translate to the position of the npc and rotate
AffineTransform npcTransform = AffineTransform.getRotateInstance(Math.toRadians(npc.angle), npc.x, npc.y);
// Translate back a few units to keep the npc rotating about its own center
// point
npcTransform.translate(-npc.halfWidth, -npc.halfHeight);
g2.setTransform(npcTransform);
// g2.draw(npc.rect); //<-- show bounding box if you want
g2.setColor(npc.outlineColor);
g2.draw(npc.visionArc);
g2.setColor(Color.BLACK);
g2.draw(npc.ellipse);
g2.setTransform(old);
}
This is my collision detection algorithim - NPC is a superclass to ninja (Shorter range, higher peripheral)
public void checkNinjas(Level level) {
for (int i = 0; i < level.ninjas.size(); i++) {
Ninja ninja = level.ninjas.get(i);
playerRect = level.player.rect;
// Check collision
if (playerRect.getBounds2D().intersects(ninja.visionArc.getBounds2D())) {
// Create an area of the object for greater precision
Area area = new Area(playerRect);
area.intersect(new Area(ninja.visionArc));
// After checking if the area intersects a second time make the NPC "See" the player
if (!area.isEmpty()) {
ninja.seesPlayer = true;
}
else {
ninja.seesPlayer = false;
}
}
}
}
Can you help me correct the actual positions of the arcs for my collision detection? I have tried creating new shapes so I can have one to do math on and one to draw to the screen but I scrapped that and am starting again from here.
[1]: https://i.stack.imgur.com/rUvTM.png
[2]: https://github.com/ShadowDraco/ArcCollisionDetection
After a few days of coding and learning and testing new ideas I came back to this program and implemented the collision detection using my original idea (ray casting) and have created the equivalent with rays!
Screenshot of the new product
Github link to the project that taught me the solution
Here's the new math
public void setRays() {
for (int i = 0; i < rays.length; i++) {
double rayStartAngleX = Math.sin(Math.toRadians((startAngle - angle) + i));
double rayStartAngleY = Math.cos(Math.toRadians((startAngle - angle) + i));
rays[i].setLine(cx, cy, cx + visionDistance * rayStartAngleX, cy + visionDistance * rayStartAngleY);
}
}
Here is a link the the program I started after I asked this question and moved on to learn more, and an image to what the new product looks like
(The original github page has been updated with a new branch :) I'm learning git hub right now too
I do not believe that using Arc2D in the way I intended is possible, however there is .setArcByTangent method, it may be possible to use that but I wasn't going to get into that. Rays are cooler.
I have created Chain shapes (static bodies) all around the perimeter of the game screen whose dimensions are:
(2f*camera.viewportWidth,2f*camera.viewportHeight).
Each chain shape has a density of 5f.
I've also created a Circle Shape (dynamic body) whose
density = 0.5f
friction = 0.25f
restitution =0.2f.
In addition, I've created a polygon shape set as a box shape (dynamic body) which has the same density, restitution and friction as the circle shape.
The world's gravity is (0,-5.8f).
All shapes render appropriately. However, the box shape just keeps falling right through the bottom chain shape which is located at the bottom of the screen. The circle shape doesn't go through, but the box shape does go through. I don't want this to happen. The size of the box shape is
(1.96f*camera.viewportWidth, 1.96f*camera.viewportHeight).
The position of this body(box shape) is set to
(0.02f*camera.viewportWidth, 0.02f*camera.viewportHeight).
I don't know why the box shape just keeps falling through and is not stopped by the bottom chain shape, just like the circle shape is stopped. Can anyone provide any insight?
Also, the reason I am trying to set up my box2d world like this is to eliminate some camera lagging movement when I use camera.translate to move around the world. My idea is to move the box shape by applying linear velocities to its body. Please any thoughts would be appreciated.
You probably did not set filtering on those fixtures. Take a look at the following example
FixtureDef f = new FixtureDef();
f.density = density;
f.friction = friction;
f.restitution = restitution;
now if you want Box2D to handle collisions for you, you should tell it which fiture will collide with which and you are doing it by assigning categoryBits and maskBits like this:
f.filter.categoryBits = categoryBits; //who am I?
f.filter.maskBits = maskBits; //with who I will collide?
Bits should be short type and power of two - you cannot have two bits defined as same value. You can set many maskBits by using logical sum ('|' operator).
public static final short BIT_F1 = 2;
public static final short BIT_F2 = 8;
public static final short BIT_F3 = 32;
FixtureDef f1, f2, f3;
//creating f1, f2, f3...
f1.filter.categoryBits = BIT_F1;
f2.filter.categoryBits = BIT_F2;
f3.filter.categoryBits = BIT_F3;
f1.filter.maskBits = BIT_F1; //now every body having f1 fixture will collide with another f1
f2.filter.maskBits = BIT_F1; //f2 will collide with f1
f3.filter.maskBits = (short)(BIT_F1 | BIT_F2 | BIT_F3); //f3 collides with everything
Please notice that if you want fixture to avoid any collision you can it to be a sensor like:
f.isSensor = sensor;
Here is link to official documentation although it is not very helpful to be honest.
I want to get all the Box2D bodies in a certain radius.
That for afaik I need to use a QueryCallback. I try it like this:
private def getBodies(position: Vector2, radius: Float): LibgdxArray[Body] =
{
val lowerBound = new Vector2(position - radius)
val upperBound = new Vector2(position + radius)
val bodies: LibgdxArray[Body] = new LibgdxArray[Body]
val callback: QueryCallback = new QueryCallback
{
override def reportFixture(fixture: Fixture): Boolean =
{
bodies.add(fixture.getBody)
true
}
}
world.queryAABB(callback, lowerBound.x, lowerBound.y, upperBound.x, upperBound.y)
bodies
}
My problem is, that I always get an empty LibgdArray. Do I need to do something more than just call queryAABB on the world? Is my code not correct?
edit: Okay, I don't understand this at all: I set the radius to a bigger value. I can remove static bodies, but the dynamic bodies do not care at all :(
I add them like this:
if(Gdx.input.isButtonPressed(Buttons.RIGHT))
{
val mousePos: Vector3 = new Vector3(Gdx.input.getX, Gdx.input.getY, 0)
val worldPos = camera.unproject(mousePos)
val ballShape: CircleShape = ShapeFactory.circle(new Vector2(worldPos.x, worldPos.y), 0.5f)
val ballDef: BodyDef = new BodyDef
ballDef.`type` = BodyType.DynamicBody
val ballBody = world.createBody(ballDef)
ballBody.createFixture(ballShape, 0.5f)
ballShape.dispose
}
Potential Solution
If you want to find all objects within a given area, one method is to first find the size of the smallest object in the world, then iteratively test all points in that area from your start point to your end point (assuming a rectangular area), with the iteration step being equal to half of the smallest size found (to avoid accidentally skipping over them).
Example code:
Finding the smallest area iteration step:
public static float smallestStep(World world) {
//Start min size at max float value
Vector2 minSize = Float.MAX_VALUE;
//Gets bodies in world
Array<Body> bodies = new Array<Body>();
world.getBodies(bodies);
//Iterate over every body
for(Body body: bodies) {
//Iterate over each fixture for each body
for(Fixture fixture : body.getFixtureList()) {
//Get fixture shape
Shape shape = fixture.getShape();
//Check if shape's "radius" is smaller than current min value
if(shape.getRadius() < minSize) {
//Updates min size
minSize = shape.getRadius();
}
}
}
//Returns half of the min found value
return minValue / 2;
}
Testing if a fixture contains a given point is the next step, by iterating over all fixtures of a body and adding the body to a list if one of the fixtures of the body contains the point using the fixture.testPoint() method, you should be able to generate a list of all of the bodies that overlap the given point.
Edit 2: http://youtu.be/KiCzUZ69gpA - as you can see in this video, the shaking effect is amplified when I also render some text for each body. Observe how the ground body (blue) shakes violently when each body has some text rendered near it, and how it does not when text rendering is commented out. This has to be connected!
Edit: I've made two important additions to the original question: I've added my rendering functions, and the camera (translation) methods, and I think that the error is actually there, not in JBox2D.
I'm trying to simulate and render a lot of random bodies (2-20) connected with RevoluteJoints. One body can be connected to multiple others, and there are no separate constructions, i.e. all the bodies are interconnected.
However, when watching the live rendering, it is very shaky and unstable. By that I mean that bodies' positions (or maybe angles) seem to be randomly fluctuating for no apparent reason, making the simulation look unstable.
Here's a video of what I'm observing:
http://youtu.be/xql-ypso1ZU
Notice the middle square and the rotating rectangle. The middle square is shifting its position back and forth slightly at seemingly random intervals, and the rotating rectangle is very jittery (take a look at the point it is rotating about).
What could this effect be due? Is it some known issue with (J)Box2D, or is it an issue with my rendering system? I think that I have somehow misconfigured the physics engine, but also some floating point math in the rendering system could be the culprit.
Here's how I'm creating the bodies and the joints:
private Body setPart(Part part) {
// body definition
BodyDef bd = new BodyDef();
bd.position.set(0f, -10f);
bd.angle = 0f;
bd.type = BodyType.DYNAMIC;
// define shape of the body.
PolygonShape Shape = new PolygonShape();
Shape.setAsBox(part.width / 2, part.height / 2);
// define fixture of the body.
FixtureDef fd = new FixtureDef();
Filter filter = new Filter();
filter.groupIndex = -1;
fd.filter = filter;
fd.shape = Shape;
fd.density = 0.5f;
fd.friction = 0.3f;
fd.restitution = 0.5f;
// create the body and add fixture to it
Body body = world.createBody(bd);
body.createFixture(fd);
body.setUserData(new PartUserData());
return body;
}
private void setJoint(PartJoint partJoint) {
Body bodyOne = partToBody.get(partJoint.partOne);
Body bodyTwo = partToBody.get(partJoint.partTwo);
RevoluteJointDef jointDef = new RevoluteJointDef();
jointDef.bodyA = bodyOne;
jointDef.bodyB = bodyTwo;
jointDef.localAnchorA = partJoint.partOne
.getAnchor(partJoint.percentOne);
jointDef.localAnchorB = partJoint.partTwo
.getAnchor(partJoint.percentTwo);
// rotation
jointDef.lowerAngle = GeomUtil.circle(partJoint.rotateFrom);
jointDef.upperAngle = GeomUtil.circle(partJoint.rotateTo);
jointDef.enableLimit = true;
jointDef.maxMotorTorque = 10.0f; // TODO limit maximum torque
jointDef.motorSpeed = GeomUtil.circle(partJoint.angularVelocity);
jointDef.enableMotor = true;
world.createJoint(jointDef);
}
The time step is 0.01f.
Here is how I draw bodies:
private void drawBody(Body body) {
// setup the transforms
Vector position = camera.translate(body.getPosition());
currentGraphics.translate(position.x, position.y);
currentGraphics.rotate(body.getAngle());
// do the actual rendering
for (Fixture fixture = body.getFixtureList(); fixture != null; fixture = fixture
.getNext()) {
PolygonShape shape = (PolygonShape) fixture.getShape();
if (body.getUserData() instanceof PartUserData) {
fillShape(shape, partFillColor);
currentGraphics.setStroke(partOutlineStroke);
outlineShape(shape, partOutlineColor);
} else {
fillShape(shape, groundFillColor);
outlineShape(shape, groundOutlineColor);
}
}
// clean up
currentGraphics.rotate(-body.getAngle());
currentGraphics.translate(-position.x, -position.y);
currentGraphics.setColor(defaultColor);
currentGraphics.setStroke(defaultStroke);
}
I think that the issue might be the way I'm handling rendering of all the bodies.
This is the algorithm for each body:
1. Translate the Graphics2D object to its position
2. Rotate it by body.getAngle()
3. Render the body
4. Rotate the graphics back
5. Translate the graphics back
Could it be that amongst all these transforms something goes wrong?
When I removed the calls to camera's methods, the effect seems to have been reduced. These are the relevant camera methods:
public Vector translate(Vec2 worldPosition) {
Vector point = new Vector();
point.x = (int) (worldPosition.x * pixelsPerMeter) - position.x;
point.y = (int) (worldPosition.y * pixelsPerMeter) - position.y;
point.x = (int) (point.x * zoom);
point.y = (int) (point.y * zoom);
point.x += renderer.getWidth() / 2;
point.y += renderer.getHeight() / 2;
return point;
}
public Vector translateRelative(Vec2 worldPosition) {
Vector point = new Vector();
point.x = (int) (worldPosition.x * pixelsPerMeter);
point.y = (int) (worldPosition.y * pixelsPerMeter);
point.x = (int) (point.x * zoom);
point.y = (int) (point.y * zoom);
return point;
}
But what part of them would cause an issue?
tl;dr: I've found a solution, but haven't identified the exact problem. Quite sure it's with my translation methods.
It seems that I have identified the scope of the problem and the solution, but I am still not sure what exactly is causing this behavior.
In those translation formulas I posted in the question, all JBox2D vectors are multiplied by a scale called pixelsPerMeter. When I set this scale to a low value, the shaking effect occurs (it's also important to note that there is another factor as well, called zoom, which is usually greater for a lower pixelsPerMeter).
So, it could be that when multiplying by a relatively low pixelsPerMeter, I have to multiply with a higher zoom factor, and since I'm converting to ints in both steps, there could be some errors in the floating point math or something. Please see the translation methods I've posted in the question.
Here's a video that demonstrates this: (to be uploaded)
Notice that when I set the pixelsPerMeter to 250, shaking seems to be gone, while when I set it to 25, it's quite visible.
Your solution was the correct one. You are not supposed to use pixel units for the Box2D physics engine :)
http://box2d.org/2011/12/pixels/
and
https://code.google.com/p/box2d/wiki/FAQ#How_do_I_convert_pixels_to_meters?
I wish to determine the 2D screen coordinates (x,y) of points in 3D space (x,y,z).
The points I wish to project are real-world points represented by GPS coordinates and elevation above sea level.
For example:
Point (Lat:49.291882, Long:-123.131676, Height: 14m)
The camera position and height can also be determined as a x,y,z point. I also have the heading of the camera (compass degrees), its degree of tilt (above/below horizon) and the roll (around the z axis).
I have no experience of 3D programming, therefore, I have read around the subject of perspective projection and learnt that it requires knowledge of matrices, transformations etc - all of which completely confuse me at present.
I have been told that OpenGL may be of use to construct a 3D model of the real-world points, set up the camera orientation and retrieve the 2D coordinates of the 3D points.
However, I am not sure if using OpenGL is the best solution to this problem and even if it is I have no idea how to create models, set up cameras etc
Could someone suggest the best method to solve my problem? If OpenGL is a feasible solution i'd have to use OpenGL ES if that makes any difference. Oh and whatever solution I choose it must execute quickly.
Here's a very general answer. Say the camera's at (Xc, Yc, Zc) and the point you want to project is P = (X, Y, Z). The distance from the camera to the 2D plane onto which you are projecting is F (so the equation of the plane is Z-Zc=F). The 2D coordinates of P projected onto the plane are (X', Y').
Then, very simply:
X' = ((X - Xc) * (F/Z)) + Xc
Y' = ((Y - Yc) * (F/Z)) + Yc
If your camera is the origin, then this simplifies to:
X' = X * (F/Z)
Y' = Y * (F/Z)
You do indeed need a perspective projection and matrix operations greatly simplify doing so. I assume you are already aware that your spherical coordinates must be transformed to Cartesian coordinates for these calculations.
Using OpenGL would likely save you a lot of work over rolling your own software rasterizer. So, I would advise trying it first. You can prototype your system on a PC since OpenGL ES is not too different as long as you keep it simple.
If just need to compute coordinates of some points, you should only need some algebra skills, not 3D programming with openGL.
Moreover openGL does not deal with Geographic coordinates
First get some doc about WGS84 and geodesic coordinates, you have first to convert your GPS data into a cartesian frame ( for instance the earth centric cartesian frame in which is defined the WGS84 ellipsoid ).
Then the computations with matrix can take place.
The chain of transformations is roughly :
WGS84
earth centric coordinates
some local frame
camera frame
2D projection
For the first conversion see this
The last involves a projection matrix
The others are only coordinates rotations and translation.
The "some local frame" is the local cartesian frame with origin as your camera location
tangent to the ellipsoid.
I'd recommend "Mathematics for 3D Game Programming and Computer Graphics" by Eric Lengyel. It covers matrices, transformations, the view frustum, perspective projection and more.
There is also a good chapter in The OpenGL Programming Guide (red book) on viewing transformations and setting up a camera (including how to use gluLookAt).
If you aren't interested in displaying the 3D scene and are limited to using OpenGL ES then it may be better to just write your own code to do the mapping from 3D to 2D window coords. As a starting point you could download Mesa 3D, an open source implementation of OpenGL, to see how they implement gluPerspective (to set a projection matrix), gluLookAt (to set a camera transformation) and gluProject (to project a 3D point to 2D window coords).
return [((fol/v[2])*v[0]+x),((fol/v[2])*v[1]+y)];
Point at [0,0,1] will be x=0 and y=0, unless you add center screen xy - it's not camera xy. fol is focal length, derived from fov angle and screen width - how high is the triangle (tangent). This method will not match three.js perspective matrix, which is why am I looking for that.
I should not be looking for it. I matched xy on openGL, perfectly like super glue! But I cannot get it to work right in java. THAT Perfect match follows.
var pmat = [0,0,0,0,0,0,0,0,0,0,
(farclip + nearclip) / (nearclip - farclip),-1,0,0,
2*farclip*nearclip / (nearclip - farclip),0 ];
void setpmat() {
double fl; // = tan(dtor(90-fovx/aspect/2)); /// UNIT focal length
fl = 1/tan(dtor(fov/Aspect/2)); /// same number
pmat[0] = fl/Aspect;
pmat[5] = fl;
}
void fovmat(double v[],double p[]) {
int cx = (int)(_Width/2),cy = (int)(_Height/2);
double pnt2[4], pnt[4] = { 0,0,0,1 } ;
COPYVECTOR(pnt,p);NORMALIZE(pnt);
popmatrix4(pnt2,pmat,pnt);
COPYVECTOR(v,pnt2);
v[0] *= -cx; v[1] *= -cy;
v[0] += cx; v[1] += cy;
} // world to screen matrix
void w2sm(int xy[],double p[]) {
double v[3]; fovmat(v,p);
xy[0] = (int)v[0];
xy[1] = (int)v[1];
}
I have one more way to match three.js xy, til I get the matrix working, just one condition. must run at Aspect of 2
function w2s(fol,v,x,y) {
var a = width / height;
var b = height/width ;
/// b = .5 // a = 2
var f = 1/Math.tan(dtor(_fov/a)) * x * b;
return [intr((f/v[2])*v[0]+x),intr((f/v[2])*v[1]+y)];
}
Use it with the inverted camera matrix, you will need invert_matrix().
v = orbital(i);
v = subv(v,campos);
v3 = popmatrix(wmatrix,v); //inverted mat
if (v3[2] > 0) {
xy = w2s(flen,v3,cx,cy);
Finally here it is, (everyone ought to know by now), the no-matrix match, any aspect.
function angle2fol(deg,centerx) {
var b = width / height;
var a = dtor(90 - (clamp(deg,0.0001,174.0) / 2));
return asa_sin(PI_5,centerx,a) / b;
}
function asa_sin(a,s,b) {
return Math.sin(b) * (s / Math.sin(PI-(a+b)));
} // ASA solve opposing side of angle2 (b)
function w2s(fol,v,x,y) {
return [intr((fol/v[2])*v[0]+x),intr((fol/v[2])*v[1]+y)];
}
Updated the image for the proof. Input _fov gets you 1.5 that, "approximately." To see the FOV readout correctly, redo the triangle with the new focal length.
function afov(deg,centerx) {
var f = angle2fol(deg,centerx);
return rtod(2 * sss_cos(f,centerx,sas_cos(f,PI_5,centerx)));
}
function sas_cos(s,a,ss) {
return Math.sqrt((Math.pow(s,2)+Math.pow(ss,2))-(2*s*ss*Math.cos(a)));
} // Side Angle Side - solve length of missing side
function sss_cos(a,b,c) {
with (Math) {
return acos((pow(a,2)+pow(c,2)-pow(b,2))/(2*a*c));
}
} // SSS solve angle opposite side2 (b)
Star library confirmed the perspective, then possible to measure the VIEW! http://innerbeing.epizy.com/cwebgl/perspective.jpg
I can explain the 90 deg correction to moon's north pole in one word precession. So what is the current up vector. pnt? radec?
function ininorths() {
if (0) {
var c = ctime;
var v = LunarPos(jdm(c));
c += secday();
var vv = LunarPos(jdm(c));
vv = crossprod(v,vv);
v = eyeradec(vv);
echo(v,vv);
v = [266.86-90,65.64]; //old
}
var v = [282.6425,65.8873]; /// new.
// ...
}
I have yet to explain the TWO sets of vectors: Three.milkyway.matrix and the 3D to 2D drawing. They ARE:
function drawmilkyway() {
var v2 = radec2pos(dtor(192.8595), dtor(27.1283),75000000);
// gcenter 266.4168 -29.0078
var v3 = radec2pos(dtor(266.4168), dtor(-29.0078),75000000);
// ...
}
function initmwmat() {
var r,u,e;
e = radec2pos(dtor(156.35), dtor(12.7),1);
u = radec2pos(dtor(60.1533), dtor(25.5935),1);
r = normaliz(crossprod(u,e));
u = normaliz(crossprod(e,r));
e = normaliz(crossprod(r,u));
var m = MilkyWayMatrix;
m[0]=r[0];m[1]=r[1];m[2]=r[2];m[3]=0.0;
m[4]=u[0];m[5]=u[1];m[6]=u[2];m[7]=0.0;
m[8]=e[0];m[9]=e[1];m[10]=e[2];m[11]=0.0;
m[12]=0.0;m[13]=0.0;m[14]=0.0;m[15]=1.0;
}
/// draw vectors and matrix were the same in C !
void initmwmat(double m[16]) {
double r[3], u[3], e[3];
radec2pos(e,dtor(192.8595), dtor(27.1283),1); //up
radec2pos(u,dtor(266.4051), dtor(-28.9362),-1); //eye
}