I solved my problem with this post, thanks.

This post might be of interest, here's similar functionality tailored for iOS: http://ar.qualcomm.at/node/2000738

- Kim

Hi,

I'm trying to adapt it to iPhone, or find something similar. When I use this code, there is an offset.

Here is the function adapted :

- (CGPoint) cameraPointToScreenPoint:(QCAR::Vec2F)cameraPoint
{
    QCAR::VideoMode videoMode = QCAR::CameraDevice::getInstance().getVideoMode(QCAR::CameraDevice::MODE_DEFAULT);
    QCAR::VideoBackgroundConfig config = QCAR::Renderer::getInstance().getVideoBackgroundConfig();
    
    int xOffset = ((int) self.frame.size.width - config.mSize.data[0]) / 2.0f + config.mPosition.data[0];
    int yOffset = ((int) self.frame.size.height - config.mSize.data[1]) / 2.0f - config.mPosition.data[1];
    
    if ([[UIApplication sharedApplication] statusBarOrientation] == UIInterfaceOrientationPortrait)
    {
        // camera image is rotated 90 degrees
        int rotatedX = videoMode.mHeight - cameraPoint.data[1];
        int rotatedY = cameraPoint.data[0];
        
        return CGPointMake(rotatedX * config.mSize.data[0] / (float) videoMode.mHeight + xOffset,
                           rotatedY * config.mSize.data[1] / (float) videoMode.mWidth + yOffset);
    }
    else
    {
        return CGPointMake(cameraPoint.data[0] * config.mSize.data[0] / (float) videoMode.mWidth + xOffset,
                           cameraPoint.data[1] * config.mSize.data[1] / (float) videoMode.mHeight + yOffset);
    }
}

Is there something different on iPhone ?

Sorry if this is an Androïd restricted area ;)

Yep, that's exactly right, use the half width and height to find the corners.

QCAR::Vec3F(-x,y,0) would get you the upper left corner.

- Kim

ultrah wrote:

sure, the code I posted above works just fine.

xyPoint.data contains the coordinates. I think if you modify QCAR::Vec3F(0,0,0) you could somehow get the target's corners

I could post some more code that tranfers the data to java if you want.

So I'm guessing QCAR::Vec3F(0,0,0) means the center of the target. So the xyPoint would be the x,y coords for the target center?

To get the left corner would it be this?
QCAR::Vec2F target_size = target->getSize();
GLfloat x= target_size.data[0]/2;
GLfloat y = target_size.data[1]/2;

QCAR::Vec3F(-x,y,0); ???

sure, the code I posted above works just fine.

xyPoint.data contains the coordinates. I think if you modify QCAR::Vec3F(0,0,0) you could somehow get the target's corners

I could post some more code that tranfers the data to java if you want.

did you figure this out? How to get x,y coords of target?

thank you for your answer ksiva.

first, my knowledge in openGL is very limited, but I think the standart openGL transformations are not available in ES 2.0:
"OpenGL ES 2.0 does not support the fixed function transformation and fragment pipeline of OpenGL ES 1.x."

if I understand your second suggestions correctly, this code should get me the screen point.

[CODE" />
const QCAR::Tracker& tracker = QCAR::Tracker::getInstance();
const QCAR::CameraCalibration& cameraCalibration = tracker.getCameraCalibration();

QCAR::Vec2F cameraPoint = QCAR::Tool::projectPoint(cameraCalibration,trackable->getPose(), QCAR::Vec3F(0,0,0));
QCAR::Vec2F xyPoint = cameraPointToScreenPoint(cameraPoint);
[/CODE" />

*edit* found my error, code updated. coordinates look good, thanks !

Or, you can use our Tool:: projectPoint method to get the point into camera space, and then use the code snippets in this thread to take the camera space point to screen space:

http://ar.qualcomm.at/node/2000570

Note that from the pose's perspective (0, 0, 0) is the center of the trackable.

- Kim

ok, so the position vector seems to be stored in [12] to [14], but x and y seem to be relative to z and z is not linear.

Is there a util to normalize x and y ?

ok, maybe I should eloborate a bit...

when a frame is rendered and a trackable is detected I can its position using:
trackable->getPose());

which gets me an undocumented '3x4 QCAR pose matrix'.

and can transform it to an (also undocumented)' QCAR::Matrix44F':
QCAR::Matrix44F modelViewMatrix = QCAR::Tool::convertPose2GLMatrix(trackable->getPose());

So how can I derive the user space position from this ?