使用陀螺仪数据旋转虚拟相机-iOS上的OpenGL ES

我创建了一个基于Vuforia平台的增强现实应用程序。我正在对其进行修改，以便在目标丢失时，系统将使用目标的最后一个已知位置，以及CoreMotion提供的设备方向数据，将对象保持在正确的位置

我需要最后一位的帮助-整合CoreMotion数据。我认为最好的方法是基于陀螺输入旋转虚拟摄像机，但我不是OpenGL ES专家。有人能告诉我们最好的方法吗？我知道如何获取设备方向数据，这是OpenGL和矩阵代数，我需要一些指导

下面是我的renderFrame方法

-(void)renderFrameQCAR {
    [self setFramebuffer];

    // Clear colour and depth buffers
    glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);

    // Render video background and retrieve tracking state
    QCAR::State state = QCAR::Renderer::getInstance().begin();
    QCAR::Renderer::getInstance().drawVideoBackground();

    glEnable(GL_DEPTH_TEST);
    glEnable(GL_CULL_FACE);


    // Check if any trackables are visible.
    int numberOfTrackables = state.getNumActiveTrackables();
    QCAR::Matrix44F modelViewMatrix;


    // Skip rendering if there is nothing to render.
    if (numberOfTrackables > 0 || hasPickedUpTrackablePreviously == YES) {

        // If there are none, but one was picked up in the past, use the last pose matrix.
        if (numberOfTrackables == 0 && hasPickedUpTrackablePreviously == YES) {
            modelViewMatrix = trackablePoseMatrix;
        }
        else {
            // Get the trackable
            const QCAR::Trackable* trackable = state.getActiveTrackable(0);
            modelViewMatrix = QCAR::Tool::convertPose2GLMatrix(trackable->getPose());

            // Store these variables for use later.
            trackablePoseMatrix = modelViewMatrix;
            hasPickedUpTrackablePreviously = YES;
        }


        // Fetch the 3D object to render.
        Object3D *obj3D;

        if (currentlyChangingTextures == YES || useDummyModel == YES) {
            obj3D = dummyObject;
        }
        else {
            obj3D = [objects3D objectAtIndex:0];
        }


        // Render using the appropriate version of OpenGL
        // OpenGL 2
        QCAR::Matrix44F modelViewProjection;

        // Apply usual transformations here
        ShaderUtils::translatePoseMatrix(sideToSideFloat, forwardBackFloat, 0.0f, &modelViewMatrix.data[0]);
        ShaderUtils::scalePoseMatrix(kObjectScale * sizeFloat, kObjectScale * sizeFloat, kObjectScale * sizeFloat, &modelViewMatrix.data[0]);
        ShaderUtils::rotatePoseMatrix(0.0f + rotationAngleFloat, 0.0f, 0.0f, 1.0f, &modelViewMatrix.data[0]);


        // Apply our translation vector here based on gesture info from the buttonOverlayViewController
        QCAR::Vec3F translationFromWorldPerspective = SampleMath::Vec3FTransformNormal(translationVectorFromCamerasPerspective, inverseModelViewMatrix);

        translationFromWorldPerspective = SampleMath::Vec3FNormalize(translationFromWorldPerspective);

        theTranslation.data[0] = theTranslation.data[0] + speed*translationFromWorldPerspective.data[0];
        theTranslation.data[1] = theTranslation.data[1] + speed*translationFromWorldPerspective.data[1];
        theTranslation.data[2] = 0.0f;

        ShaderUtils::translatePoseMatrix(theTranslation.data[0], theTranslation.data[1], theTranslation.data[2], &modelViewMatrix.data[0]);

        // Update inverseModelViewMatrix
        inverseModelViewMatrix = SampleMath::Matrix44FInverse(modelViewMatrix);

        // Multiply modelview and projection matrix as usual
        ShaderUtils::multiplyMatrix(&qUtils.projectionMatrix.data[0], &modelViewMatrix.data[0], &modelViewProjection.data[0]);

        glUseProgram(shaderProgramID);

        glVertexAttribPointer(vertexHandle, 3, GL_FLOAT, GL_FALSE, 0, (const GLvoid*)obj3D.vertices);
        glVertexAttribPointer(normalHandle, 3, GL_FLOAT, GL_FALSE, 0, (const GLvoid*)obj3D.normals);
        glVertexAttribPointer(textureCoordHandle, 2, GL_FLOAT, GL_FALSE, 0, (const GLvoid*)obj3D.texCoords);

        glEnableVertexAttribArray(vertexHandle);
        glEnableVertexAttribArray(normalHandle);
        glEnableVertexAttribArray(textureCoordHandle);

        glActiveTexture(GL_TEXTURE0);
        glBindTexture(GL_TEXTURE_2D, [obj3D.texture textureID]);
        glUniformMatrix4fv(mvpMatrixHandle, 1, GL_FALSE, (const GLfloat*)&modelViewProjection.data[0]);
        glDrawArrays(GL_TRIANGLES, 0, obj3D.numVertices);

        ShaderUtils::checkGlError("EAGLView renderFrameQCAR");
    }


    // Disable these things.
    glDisable(GL_DEPTH_TEST);
    glDisable(GL_CULL_FACE);

    glDisableVertexAttribArray(vertexHandle);
    glDisableVertexAttribArray(normalHandle);
    glDisableVertexAttribArray(textureCoordHandle);

    QCAR::Renderer::getInstance().end();
    [self presentFramebuffer];
}

---

谢谢

我还没有使用Vuforia，所以我不完全理解您的代码，但我已经成功创建了一个AR应用程序，使用陀螺仪和指南针控制相机。这是我的相机矩阵代码：

GLKMatrix4 cameraMatrix = GLKMatrix4Identity;
cameraMatrix = GLKMatrix4Rotate(cameraMatrix, GLKMathDegreesToRadians((zenith-90)), 1.0f, 0.0f, 0.0f);
cameraMatrix = GLKMatrix4Rotate(cameraMatrix, GLKMathDegreesToRadians(azimuth), 0.0f, 1.0f, 0.0f);

其中： 天顶=（180-陀螺仪滚动角）[0点直线上升，180点直线下降]，和 方位角=罗盘角度（北纬0度，东经90度，南纬180度，西经270度）

在顶点着色器中，glPosition通过以下公式计算：

gl_Position = uProjectionMatrix * uCameraMatrix * vec4(position, 1.0);

我使用的是点精灵，但position保留了我的粒子在3D世界空间中的坐标。在您的情况下，我猜这是您将替换其他矩阵的地方，如modelViewMatrix等

---

小心你的矩阵乘法顺序

谢谢你的建议。我没能让它工作（我不是专家），我在这方面得到了额外的帮助。如果解决方案不太复杂，我会在这里发布。