In my glut application I'm simulating a plane with the camera. When the planes speed is low I intend to have the nose start to point towards the ground as the camera falls. My first instinct was to just change the pitch until it was pointed downwards at -90degrees. However I can't just change the pitch because if the plane is tilted on its side or upside down then it would note be changing direction towards the ground.
Now i'm trying to do a rough simulation of this by shifting the 'lookAt.y' downwards. To do this I am trying to get all the current camera coordinates that I use to set the camera
(eye.x, eye.y, eye.z, look.x, look.y, look.z, up.x, up.y, up.z). Then recall the set with the new modified values.
I've been working with the Camera.cpp and Camera.h to control my camera functions. They can be found here
after adding methods to get all the values, only the eye values are actually updated when various camera motions are made. I guess my question is how do I retrieve these values.
The glLoadMaxtrix call is in this function
void Camera :: setModelViewMatrix(void)
{ // load model view matrix with existing camera values
float m[16];
Vector3 eVec(eye.x, eye.y, eye.z);
m[0] = u.x; m[4] = u.y; m[8] = u.z; m[12] = -eVec.dot(u);
m[1] = v.x; m[5] = v.y; m[9] = v.z; m[13] = -eVec.dot(v);
m[2] = n.x; m[6] = n.y; m[10] = n.z; m[14] = -eVec.dot(n);
m[3] = 0; m[7] = 0; m[11] = 0; m[15] = 1.0;
look.x = u.y; look.y = v.y; look.z = n.y;
glMatrixMode(GL_MODELVIEW);
glLoadMatrixf(m);
}
Is there a way to get 'eye', 'lookAt', and 'up' values from the matrix here? Or should I do something else to get these values?
-Thanks in advance for your help
The camera class you link to is not an actual OpenGL class, but it should be simple enough to work with.
The function quoted just takes the current values of the camera object and sends them to OpenGL. If you look at the camera's set function, you can see how the program calculates the values it actually stores.
The eye value is stored directly. The lookAt value is just the value of (eye - n), by vector math. The up value is the hardest, but if I remember my vector math correctly, I believe that up = (n cross u).
Related
I am making a game based on the game AZ on the website Y8, and I am having problems with tile collisions.
the player moves basically by giving it speed when up is pressed, then rotating left or right.
direction = image_angle;
if(keyForward)
{
speed = 2;
}
else speed = 0;
// rotate
if(keyRotateLeft)
{
image_angle = image_angle + 5;
}
if(keyRotateRight)
{
image_angle = image_angle - 5;
}
then I said when the player collides with the tile speed = 0. But the player gets stuck and can't move anymore. is there a better way to do this.
A simple approach would be as following:
Attempt to rotate
Check if you are now stuck in a wall
If you are, undo the rotation.
A more advanced approach would be to attempt pushing the player out of solids while rotating.
Alternatively, you may be able to get away with giving the player a circular mask and not rotating the actual mask (using a user-defined variable instead of image_angle).
There are other posts on Stack Overflow on pinch zooming, but I haven't found any helpful ones for OpenGL that do what I'm looking for. I am currently using the orthoM function to change the camera position and to do scaling in OpenGL. I have gotten the camera to move around, and have gotten pinch zooming to work, but the zooming always zooms into the center of the OpenGL surface view coordinate system at 0,0. After trying different things, I haven't found a way yet that allows the camera to move around, while also allowing pinch zooming to the user's touch point (as an example, the touch controls in Clash of Clans is similar to what I am trying to make).
(The method I'm currently using to get the scale value is based on this post.)
My first attempt:
// mX and mY are the movement offsets based on the user's touch movements,
// and can be positive or negative
Matrix.orthoM(mProjectionMatrix, 0, ((-WIDTH/2f)+mX)*scale, ((WIDTH/2f)+mX)*scale,
((-HEIGHT/2f)+mY)*scale, ((HEIGHT/2f)+mY)*scale, 1f, 2f);
In the above code, I realize that the camera moves towards the coordinate 0,0 because as scale gets increasingly smaller, the values for the camera edges decrease towards 0. So although the zoom goes towards the coordinate system center, the movement of the camera moves at the right speeds at any scale level.
So, I then edited the code to this:
Matrix.orthoM(mProjectionMatrix, 0, (-WIDTH/2f)*scale+mX, (WIDTH/2f)*scale+mX,
(-HEIGHT/2f)*scale+mY, (HEIGHT/2f)*scale+mY, 1f, 2f);
The edited code now makes the zoom go toward the center of the screen no matter where in the surface view coordinate system the camera is (although that isn't the full goal), but the camera movement is off, as the offset isn't adjusted for the different scale levels.
I'm still working to find a solution myself, but if anyone has any advice or ideas on how this could be implemented, I would be glad to hear.
Note, I don't think it matters, but I'm doing this in Android and using Java.
EDIT:
Since I first posted this question, I have made some changes to my code. I found this post, which explains the logic of how to pan the camera to the correct position based on the scale, so that the zoompoint remains in the same position.
My updated attempt:
// Only do the following if-block if two fingers are on the screen
if (zooming) {
// midPoint is a PointF object that stores the coordinate of the midpoint between
//two fingers
float scaleChange = scale - prevScale; // scale is the same as in my previous code
float offsetX = -(midPoint.x*scaleChange);
float offsetY = -(midPoint.y*scaleChange);
cameraPos.x += offsetX;
cameraPos.y += offsetY;
}
// cameraPos is a PointF object that stores the coordinate at the center of the screen,
// and replaces the previous values mX and mY
left = cameraPos.x-(WIDTH/2f)*scale;
right = cameraPos.x+(WIDTH/2f)*scale;
bottom = cameraPos.y-(HEIGHT/2f)*scale;
top = cameraPos.y+(HEIGHT/2f)*scale;
Matrix.orthoM(mProjectionMatrix, 0, left, right, bottom, top, 1f, 2f);
The code does work quite a bit better now, but it still isn't completely accurate. I tested how the code worked when panning was disabled, and the zooming worked sort of better. However, when the panning is enabled, the zooming doesn't focus in on the zoompoint at all.
I finally found a solution while working on another project, so I'll post (in simplest form possible) what worked for me in case this could help anyone by chance.
final float currentPointersDistance = this.calculateDistance(pointer1CurrentX, pointer1CurrentY, pointer2CurrentX, pointer2CurrentY);
final float zoomFactorMultiplier = currentPointersDistance/initialPointerDistance; //> Get an initial distance between two pointers before calling this
final float newZoomFactor = previousZoomFactor*zoomFactorMultiplier;
final float zoomFactorChange = newZoomFactor-previousZoomFactor; //> previousZoomFactor is the current value of the zoom
//> The x and y values of the variables are in scene coordinate form (not surface)
final float distanceFromCenterToMidpointX = camera.getCenterX()-currentPointersMidpointX;
final float distanceFromCenterToMidpointY = camera.getCenterY()-currentPointersMidpointY;
final float offsetX = -(distanceFromCenterToMidpointX*zoomFactorChange/newZoomFactor);
final float offsetY = -(distanceFromCenterToMidpointY*zoomFactorChange/newZoomFactor);
camera.setZoomFactor(newZoomFactor);
camera.translate(offsetX, offsetY);
initialPointerDistance = currentPointersDistance; //> Make sure to do this
Method used to calculate the distance between two pointers:
public float calculateDistance(float pX1, float pY1, float pX2, float pY2) {
float x = pX2-pX1;
float y = pY2-pY1;
return (float)Math.sqrt((x*x)+(y*y));
}
Camera class methods used above:
public float getXMin() {
return centerX-((centerX-xMin)/zoomFactor);
}
public float getYMin() {
return centerY-((centerY-yMin)/zoomFactor);
}
public float getXMax() {
return centerX+((xMax-centerX)/zoomFactor);
}
public float getYMax() {
return centerY+((yMax-centerY)/zoomFactor);
}
public void setZoomFactor(float pZoomFactor) {
zoomFactor = pZoomFactor;
}
public void translate(float pX, float pY) {
xMin += pX;
yMin += pY;
xMax += pX;
yMax += pY;
}
The orthoM() function is called like the following:
Matrix.orthoM(projectionMatrix, 0, camera.getXMin(), camera.getXMax(), camera.getYMin(), camera.getYMax(), near, far);
I am struggling in solving this problem.
On my scene, I have a camera which looks at the center of mass of an object. I have a some buttons that enable to set camera position on particular view (front view, back view,...) along a invisible sphere that surroung the object (constant radius).
When I click on the button, i would like the camera to move from its start position to the end position along the sphere surface. When camera moves I would like it to keep fixing center of mass of the object.
Has anyone have a clue on how to achieve this?
Thanks for help!
If you are happy/prefer to use basic trigonometry then in your initialisation section you could do this:
var cameraAngle = 0;
var orbitRange = 100;
var orbitSpeed = 2 * Math.PI/180;
var desiredAngle = 90 * Math.PI/180;
...
camera.position.set(orbitRange,0,0);
camera.lookAt(myObject.position);
Then in your render/animate section you could do this:
if (cameraAngle == desiredAngle) { orbitSpeed = 0; }
else {
cameraAngle += orbitSpeed;
camera.position.x = Math.cos(cameraAngle) * orbitRange;
camera.position.y = Math.sin(cameraAngle) * orbitRange;
}
Of course, your buttons would modify what the desiredAngle was (0°, 90°, 180° or 270° presumably), you need to rotate around the correct plane (I am rotating around the XY plane above), and you can play with the orbitRange and orbitSpeed until you hare happy.
You can also modify orbitSpeed as it moves along the orbit path, speeding up and slowing down at various cameraAngles for a smoother ride. This process is called 'tweening' and you could search on 'tween' or 'tweening' if you want to know more. I think Three.js has tweening support but have never looked into it.
Oh, also remember to set your camera's far property to be greater than orbitRadius or you will only see the front half of your object and, depending on what it is, that might look weird.
I am implementing a Forward Renderer with DirectX 10. I want it to handle an unlimited amount of lights so I can later compare its performance with a Deferred Renderer. So basically the algorithm I am using is: for every object, for every light -> set light, draw object. Using additive blending, I render the object for each light summing the contribution of every light on it. Everything works using an additive blending and disabling depth writes. The problem I have is that, using this simple approach, different object get blended together (because depth writes are disabled), while I just want a single object to be blended with the different light contribution's on it but still to obscure other objects behind it. How can I do this? Is a Z pre-pass the solution? Any suggestion will be very appreciated. Thanks.
This are the blending and depth/stencil states I use in my HLSL shader:
DepthStencilState NoDepthWritesDSS
{
DepthEnable = true;
DepthWriteMask = Zero;
StencilEnable = true;
StencilReadMask = 0xff;
StencilWriteMask = 0xff;
FrontFaceStencilFunc = Always;
FrontFaceStencilPass = Incr;
FrontFaceStencilFail = Keep;
BackFaceStencilFunc = Always;
BackFaceStencilPass = Incr;
BackFaceStencilFail = Keep;
};
BlendState BlendingAddBS
{
AlphaToCoverageEnable = false;
BlendEnable[0] = true;
SrcBlend = ONE;
DestBlend = ONE;
BlendOp = ADD;
SrcBlendAlpha = ZERO;
DestBlendAlpha = ZERO;
BlendOpAlpha = ADD;
RenderTargetWriteMask[0] = 0x0F;
};
There's several options to handle multiple lights, if you want to implement it using multipass a depth pre pass is your best option (then you do draw again using LESS_EQUAL comparison on your depth state).
This approach will most likely be quite unefficient on a high number of lights/objects tho.
I recommend this article which explains how to render several lights, it has different interesting implementations. The compute tile will not work in directx10, but the geometry sprite version can be easily ported (I have a dx9 version of it)
If you still want forward rendering, there's also the light indexed technique, implementation example here
I'm searching for a program which detects the border of a image,
for example I have a square and the program detects the X/Y-Coords
Example:
alt text http://img709.imageshack.us/img709/1341/22444641.png
This is a very simple edge detector. It is suitable for binary images. It just calculates the differences between horizontal and vertical pixels like image.pos[1,1] = image.pos[1,1] - image.pos[1,2] and the same for vertical differences. Bear in mind that you also need to normalize it in the range of values 0..255.
But! if you just need a program, use Adobe Photoshop.
Code written in C#.
public void SimpleEdgeDetection()
{
BitmapData data = Util.SetImageToProcess(image);
if (image.PixelFormat != PixelFormat.Format8bppIndexed)
return;
unsafe
{
byte* ptr1 = (byte *)data.Scan0;
byte* ptr2;
int offset = data.Stride - data.Width;
int height = data.Height - 1;
int px;
for (int y = 0; y < height; y++)
{
ptr2 = (byte*)ptr1 + data.Stride;
for (int x = 0; x < data.Width; x++, ptr1++, ptr2++)
{
px = Math.Abs(ptr1[0] - ptr1[1]) + Math.Abs(ptr1[0] - ptr2[0]);
if (px > Util.MaxGrayLevel) px = Util.MaxGrayLevel;
ptr1[0] = (byte)px;
}
ptr1 += offset;
}
}
image.UnlockBits(data);
}
Method from Util Class
static public BitmapData SetImageToProcess(Bitmap image)
{
if (image != null)
return image.LockBits(
new Rectangle(0, 0, image.Width, image.Height),
ImageLockMode.ReadWrite,
image.PixelFormat);
return null;
}
If you need more explanation or algorithm just ask with more information without being so general.
It depends what you want to do with the border, if you are looking at getting just the values of the edges of the region, use an algorithm called the Connected Components Region. You must know the value of the region prior to using the algorithm. This will navigate around the border and collect the outside region. If you are trying to detect just the outside lines get the gradient of the image and it will reveal where the lines are. To do this convolve the image with an edge detection filter such as Prewitt, Sobel, etc.
You can use any image processing library such as Opencv. which is in c++ or python.
You should look for edge detection functions such as Canny edge detection.
Of course this would require some diving into image processing.
The example image you gave should be straight forward to detect, how noisy/varied are the images going to be?
A shape recognition algorithm might help you out, providing it has a solid border of some kind, and the background colour is a solid one.
From the sounds of it, you just want a blob extraction algorithm. After that, the lowest/highest values for x/y will give you the coordinates of the corners.