Android Camera 系统框架分析

一、在android中开发人员可以做那些工作？
        应用程序开发：利用android提供的强大的sdk，开发出各种各样新颖的应用。
        系统开发：在android中Google实现了与硬件无关的所有代码，但是与硬件密切相关的硬件抽象层却没有也无法提供，对于移动设备不同的设备提供商 底层硬件是千变万化的，不可能提供统一的硬件驱动以及接口实现，只能提供标准的接口，因此硬件提供商需要自个儿开发设备驱动，
并去实现android框架提供的接口。
二、android框架中Camera系统源码分析
       在每个android手机中都有一个Camera应用程序用来实现拍照功能，不同硬件提供商可能会对这个应用程序进行改变来适合自己的UI风格，
这里仅仅分析android原生Camera应用以及框架（Android 4.0）
      原生Camera应用代码在Camera.java(android4.0\packages\apps\camera\src\com\android\camera)，这个应该算是Camera系统最上层，应用层的实现。
      下面是Camera类部分代码

public class Camera extends ActivityBase implements FocusManager.Listener,
        View.OnTouchListener, ShutterButton.OnShutterButtonListener,
        SurfaceHolder.Callback, ModePicker.OnModeChangeListener,
        FaceDetectionListener, CameraPreference.OnPreferenceChangedListener,
        LocationManager.Listener, ShutterButton.OnShutterButtonLongPressListener

      从上面可以看出，Camera在继承了很多监听接口，用来监听各种事件（对焦事件、用户触摸事件等）。这个应用时继承ActivityBase，
可以重载OnCreate、OnResume等接口，在这些接口中完成相关初始化的工作，基本就是初始化各种监听对象，以及获取相机参数等相关。
     比较关键的在 doOnResume这个函数中：

@Override
    protected void doOnResume() {
        if (mOpenCameraFail || mCameraDisabled) return;

        mPausing = false;

        mJpegPictureCallbackTime = 0;
        mZoomValue = 0;

        // Start the preview if it is not started.
        if (mCameraState == PREVIEW_STOPPED) {
            try {
                mCameraDevice = Util.openCamera(this, mCameraId);
                initializeCapabilities();
                resetExposureCompensation();
                startPreview();
                if (mFirstTimeInitialized) startFaceDetection();
            } catch (CameraHardwareException e) {
                Util.showErrorAndFinish(this, R.string.cannot_connect_camera);
                return;
            } catch (CameraDisabledException e) {
                Util.showErrorAndFinish(this, R.string.camera_disabled);
                return;
            }
        }

        if (mSurfaceHolder != null) {
            // If first time initialization is not finished, put it in the
            // message queue.
            if (!mFirstTimeInitialized) {
                mHandler.sendEmptyMessage(FIRST_TIME_INIT);
            } else {
                initializeSecondTime();
            }
        }
        keepScreenOnAwhile();

        if (mCameraState == IDLE) {
            mOnResumeTime = SystemClock.uptimeMillis();
            mHandler.sendEmptyMessageDelayed(CHECK_DISPLAY_ROTATION, 100);
        }
    }

在这个函数中看到通过这个函数获得Camera底层对象
mCameraDevice = Util.openCamera(this, mCameraId)，这里使用Util这个类，这个类的实现在
Util.java (android4.0\packages\apps\camera\src\com\android\camera)中，找到OpenCamera这个函数实现：

public static android.hardware.Camera openCamera(Activity activity, int cameraId)
            throws CameraHardwareException, CameraDisabledException {
        // Check if device policy has disabled the camera.
        DevicePolicyManager dpm = (DevicePolicyManager) activity.getSystemService(
                Context.DEVICE_POLICY_SERVICE);
        if (dpm.getCameraDisabled(null)) {
            throw new CameraDisabledException();
        }

        try {
            return CameraHolder.instance().open(cameraId);
        } catch (CameraHardwareException e) {
            // In eng build, we throw the exception so that test tool
            // can detect it and report it
            if ("eng".equals(Build.TYPE)) {
                throw new RuntimeException("openCamera failed", e);
            } else {
                throw e;
            }
        }
    }

从这个函数可以看出，android系统中对下层Camera管理，是通过一个单例模式CameraHolder来管理的，
定位到这个类的实现CameraHolder.java (android4.0\packages\apps\camera\src\com\android\camera)通过调用open函数获取一个Camera硬件设备对象，
因为Camera设备是独享设备，不能同时被两个进程占用，而整个android系统是一个多进程环境，因此需要加入一些进程间互斥同步的方法。
定位到这个类的open函数：

public synchronized android.hardware.Camera open(int cameraId)
            throws CameraHardwareException {
        Assert(mUsers == 0);
        if (mCameraDevice != null && mCameraId != cameraId) {
            mCameraDevice.release();
            mCameraDevice = null;
            mCameraId = -1;
        }
        if (mCameraDevice == null) {
            try {
                Log.v(TAG, "open camera " + cameraId);
                mCameraDevice = android.hardware.Camera.open(cameraId);
                mCameraId = cameraId;
            } catch (RuntimeException e) {
                Log.e(TAG, "fail to connect Camera", e);
                throw new CameraHardwareException(e);
            }
            mParameters = mCameraDevice.getParameters();
        } else {
            try {
                mCameraDevice.reconnect();
            } catch (IOException e) {
                Log.e(TAG, "reconnect failed.");
                throw new CameraHardwareException(e);
            }
            mCameraDevice.setParameters(mParameters);
        }
        ++mUsers;
        mHandler.removeMessages(RELEASE_CAMERA);
        mKeepBeforeTime = 0;
        return mCameraDevice;
    }

通 过android.hardware.Camera.open(cameraId)调用进入下一层封装，JNI层，这一层是java代码的最下层，对下层 CameraC++代码进行JNI封装，封装实现类在Camera.java (android4.0\frameworks\base\core\java\android\hardware) 下面是这个类的部分实现，里面定义了不少回调函数：

public class Camera {
    private static final String TAG = "Camera";

    // These match the enums in frameworks/base/include/camera/Camera.h
    private static final int CAMERA_MSG_ERROR            = 0x001;
    private static final int CAMERA_MSG_SHUTTER          = 0x002;
    private static final int CAMERA_MSG_FOCUS            = 0x004;
    private static final int CAMERA_MSG_ZOOM             = 0x008;
    private static final int CAMERA_MSG_PREVIEW_FRAME    = 0x010;
    private static final int CAMERA_MSG_VIDEO_FRAME      = 0x020;
    private static final int CAMERA_MSG_POSTVIEW_FRAME   = 0x040;
    private static final int CAMERA_MSG_RAW_IMAGE        = 0x080;
    private static final int CAMERA_MSG_COMPRESSED_IMAGE = 0x100;
    private static final int CAMERA_MSG_RAW_IMAGE_NOTIFY = 0x200;
    private static final int CAMERA_MSG_PREVIEW_METADATA = 0x400;
    private static final int CAMERA_MSG_ALL_MSGS         = 0x4FF;

    private int mNativeContext; // accessed by native methods
    private EventHandler mEventHandler;
    private ShutterCallback mShutterCallback;
    private PictureCallback mRawImageCallback;
    private PictureCallback mJpegCallback;
    private PreviewCallback mPreviewCallback;
    private PictureCallback mPostviewCallback;
    private AutoFocusCallback mAutoFocusCallback;
    private OnZoomChangeListener mZoomListener;
    private FaceDetectionListener mFaceListener;
    private ErrorCallback mErrorCallback;

定位到Open函数：
    public static Camera open(int cameraId) {
        return new Camera(cameraId);
    }
Open函数是一个静态方法,构造一个Camera对象：

Camera(int cameraId) {
        mShutterCallback = null;
        mRawImageCallback = null;
        mJpegCallback = null;
        mPreviewCallback = null;
        mPostviewCallback = null;
        mZoomListener = null;

        Looper looper;
        if ((looper = Looper.myLooper()) != null) {
            mEventHandler = new EventHandler(this, looper);
        } else if ((looper = Looper.getMainLooper()) != null) {
            mEventHandler = new EventHandler(this, looper);
        } else {
            mEventHandler = null;
        }

        native_setup(new WeakReference<Camera>(this), cameraId);
    }

在构造函数中调用native_setup方法，此方法对应于C++代码的android_hardware_Camera_native_setup方法，
实现在android_hardware_Camera.cpp (android4.0\frameworks\base\core\jni)，具体代码如下：

static void android_hardware_Camera_native_setup(JNIEnv *env, jobject thiz,
    jobject weak_this, jint cameraId)
{
    sp<Camera> camera = Camera::connect(cameraId);

    if (camera == NULL) {
        jniThrowRuntimeException(env, "Fail to connect to camera service");
        return;
    }

    // make sure camera hardware is alive
    if (camera->getStatus() != NO_ERROR) {
        jniThrowRuntimeException(env, "Camera initialization failed");
        return;
    }

    jclass clazz = env->GetObjectClass(thiz);
    if (clazz == NULL) {
        jniThrowRuntimeException(env, "Can't find android/hardware/Camera");
        return;
    }

    // We use a weak reference so the Camera object can be garbage collected.
    // The reference is only used as a proxy for callbacks.
    sp<JNICameraContext> context = new JNICameraContext(env, weak_this, clazz, camera);
    context->incStrong(thiz);
    camera->setListener(context);

    // save context in opaque field
    env->SetIntField(thiz, fields.context, (int)context.get());
}

在android_hardware_Camera_native_setup方法中调用了Camera对象的connect方法，这个Camera类的声明在Camera.h (android4.0\frameworks\base\include\camera）
定位到connect方法：

sp<Camera> Camera::connect(int cameraId)
{
    LOGV("connect");
    sp<Camera> c = new Camera();
    const sp<ICameraService>& cs = getCameraService();
    if (cs != 0) {
        c->mCamera = cs->connect(c, cameraId);
    }
    if (c->mCamera != 0) {
        c->mCamera->asBinder()->linkToDeath(c);
        c->mStatus = NO_ERROR;
    } else {
        c.clear();
    }
    return c;
}

这里以下的代码就比较关键了，涉及到Camera框架的实现机制，Camera系统使用的是Server-Client机制，Service和Client位于不同的进程中，进程间使用Binder机制进行通信，
Service端实际实现相机相关的操作，Client端通过Binder接口调用Service对应的操作。
继续分析代码，上面函数调用getCameraService方法，获得CameraService的引用，ICameraService有两个子类，BnCameraService和BpCameraService，这两个子类同时也
继承了IBinder接口，这两个子类分别实现了Binder通信的两端，Bnxxx实现ICameraService的具体功能，Bpxxx利用Binder的通信功能封装ICameraService方法，具体如下：

class ICameraService : public IInterface
{
public:
    enum {
        GET_NUMBER_OF_CAMERAS = IBinder::FIRST_CALL_TRANSACTION,
        GET_CAMERA_INFO,
        CONNECT
    };

public:
    DECLARE_META_INTERFACE(CameraService);

    virtual int32_t         getNumberOfCameras() = 0;
    virtual status_t        getCameraInfo(int cameraId,
                                          struct CameraInfo* cameraInfo) = 0;
    virtual sp<ICamera>     connect(const sp<ICameraClient>& cameraClient,
                                    int cameraId) = 0;
};

// ----------------------------------------------------------------------------

class BnCameraService: public BnInterface<ICameraService>
{
public:
    virtual status_t    onTransact( uint32_t code,
                                    const Parcel& data,
                                    Parcel* reply,
                                    uint32_t flags = 0);
};

}; // na

class BpCameraService: public BpInterface<ICameraService>
{
public:
    BpCameraService(const sp<IBinder>& impl)
        : BpInterface<ICameraService>(impl)
    {
    }

    // get number of cameras available
    virtual int32_t getNumberOfCameras()
    {
        Parcel data, reply;
        data.writeInterfaceToken(ICameraService::getInterfaceDescriptor());
        remote()->transact(BnCameraService::GET_NUMBER_OF_CAMERAS, data, &reply);
        return reply.readInt32();
    }

    // get information about a camera
    virtual status_t getCameraInfo(int cameraId,
                                   struct CameraInfo* cameraInfo) {
        Parcel data, reply;
        data.writeInterfaceToken(ICameraService::getInterfaceDescriptor());
        data.writeInt32(cameraId);
        remote()->transact(BnCameraService::GET_CAMERA_INFO, data, &reply);
        cameraInfo->facing = reply.readInt32();
        cameraInfo->orientation = reply.readInt32();
        return reply.readInt32();
    }

    // connect to camera service
    virtual sp<ICamera> connect(const sp<ICameraClient>& cameraClient, int cameraId)
    {
        Parcel data, reply;
        data.writeInterfaceToken(ICameraService::getInterfaceDescriptor());
        data.writeStrongBinder(cameraClient->asBinder());
        data.writeInt32(cameraId);
        remote()->transact(BnCameraService::CONNECT, data, &reply);
        return interface_cast<ICamera>(reply.readStrongBinder());
    }
};

IMPLEMENT_META_INTERFACE(CameraService, "android.hardware.ICameraService");

// ----------------------------------------------------------------------

status_t BnCameraService::onTransact(
    uint32_t code, const Parcel& data, Parcel* reply, uint32_t flags)
{
    switch(code) {
        case GET_NUMBER_OF_CAMERAS: {
            CHECK_INTERFACE(ICameraService, data, reply);
            reply->writeInt32(getNumberOfCameras());
            return NO_ERROR;
        } break;
        case GET_CAMERA_INFO: {
            CHECK_INTERFACE(ICameraService, data, reply);
            CameraInfo cameraInfo;
            memset(&cameraInfo, 0, sizeof(cameraInfo));
            status_t result = getCameraInfo(data.readInt32(), &cameraInfo);
            reply->writeInt32(cameraInfo.facing);
            reply->writeInt32(cameraInfo.orientation);
            reply->writeInt32(result);
            return NO_ERROR;
        } break;
        case CONNECT: {
            CHECK_INTERFACE(ICameraService, data, reply);
            sp<ICameraClient> cameraClient = interface_cast<ICameraClient>(data.readStrongBinder());
            sp<ICamera> camera = connect(cameraClient, data.readInt32());
            reply->writeStrongBinder(camera->asBinder());
            return NO_ERROR;
        } break;
        default:
            return BBinder::onTransact(code, data, reply, flags);
    }
}

// ----------------------------------------------------------------------------

}; // namespace android

下面继续分析sp<Camera> Camera::connect(int cameraId)这个方法，，定位到getCameraService这个方法

const sp<ICameraService>& Camera::getCameraService()
{
    Mutex::Autolock _l(mLock);
    if (mCameraService.get() == 0) {
        sp<IServiceManager> sm = defaultServiceManager();
        sp<IBinder> binder;
        do {
            binder = sm->getService(String16("media.camera"));
            if (binder != 0)
                break;
            LOGW("CameraService not published, waiting...");
            usleep(500000); // 0.5 s
        } while(true);
        if (mDeathNotifier == NULL) {
            mDeathNotifier = new DeathNotifier();
        }
        binder->linkToDeath(mDeathNotifier);
        mCameraService = interface_cast<ICameraService>(binder);
    }
    LOGE_IF(mCameraService==0, "no CameraService!?");
    return mCameraService;
}

定位到mCameraService = interface_cast<ICameraService>(binder); mCameraService是一个ICamerService类型，更加具体具体一点来讲应该是BpCameraService，
因为在这个类中实现了ICameraService的方法。

总结上面Binder机制，仅仅考虑分析Binder用法，对底层实现不进行深究，基本步骤如下：
1.定义进程间通信的接口比如这里的ICameraService；
2.在BnCameraService和BpCamaraService实现这个接口，这两个接口也分别继承于BnInterface和BpInterface；
3.服务端向ServiceManager注册Binder，客户端向ServiceManager获得Binder；
4.然后就可以实现双向进程间通信了；

通过getCameraService得到ICameraService引用后，调用ICameraService的connect方法获得ICamera引用，

c->mCamera = cs->connect(c, cameraId);

进一步跟进connect方法，这里就是BpCameraService类中connect方法的具体实现。

virtual sp<ICamera> connect(const sp<ICameraClient>& cameraClient, int cameraId)  
{  
    Parcel data, reply;
    data.writeInterfaceToken(ICameraService::getInterfaceDescriptor());          
    data.writeStrongBinder(cameraClient->asBinder());             
    data.writeInt32(cameraId);  
    remote()->transact(BnCameraService::CONNECT, data, &reply);  
    return interface_cast<ICamera>(reply.readStrongBinder());  
}

在这里返回的ICamera对象，实际上应该是BpCamera对象，这里使用的是匿名Binder，前面获取CameraService的 使用的有名Binder，有名Binder需要借助于ServiceManager获取Binder，而匿名Binder可以通过已经建立后的通信通道 （有名Binder）获得。以上是实现Camera框架部分，具体的实现Camera相关的方法是在ICamera相关的接口，下面是给接口的定义：

class ICamera: public IInterface  
{  
public:  
    DECLARE_META_INTERFACE(Camera);  
  
    virtual void            disconnect() = 0;  
  
    // connect new client with existing camera remote  
    virtual status_t        connect(const sp<ICameraClient>& client) = 0;  
  
    // prevent other processes from using this ICamera interface  
    virtual status_t        lock() = 0;  
  
    // allow other processes to use this ICamera interface  
    virtual status_t        unlock() = 0;  
  
    // pass the buffered Surface to the camera service  
    virtual status_t        setPreviewDisplay(const sp<Surface>& surface) = 0;  
  
    // pass the buffered ISurfaceTexture to the camera service  
    virtual status_t        setPreviewTexture(  
            const sp<ISurfaceTexture>& surfaceTexture) = 0;  
  
    // set the preview callback flag to affect how the received frames from  
    // preview are handled.  
    virtual void            setPreviewCallbackFlag(int flag) = 0;  
  
    // start preview mode, must call setPreviewDisplay first  
    virtual status_t        startPreview() = 0;  
  
    // stop preview mode  
    virtual void            stopPreview() = 0;  
  
    // get preview state  
    virtual bool            previewEnabled() = 0;  
  
    // start recording mode  
    virtual status_t        startRecording() = 0;  
  
    // stop recording mode  
    virtual void            stopRecording() = 0;  
  
    // get recording state  
    virtual bool            recordingEnabled() = 0;  
  
    // release a recording frame  
    virtual void            releaseRecordingFrame(const sp<IMemory>& mem) = 0;     
    // auto focus  
    virtual status_t        autoFocus() = 0;     
    // cancel auto focus  
    virtual status_t        cancelAutoFocus() = 0;  
    /* 
     * take a picture.       
     * @param msgType the message type an application selectively turn on/off 
     * on a photo-by-photo basis. The supported message types are: 
     * CAMERA_MSG_SHUTTER, CAMERA_MSG_RAW_IMAGE, CAMERA_MSG_COMPRESSED_IMAGE, 
     * and CAMERA_MSG_POSTVIEW_FRAME. Any other message types will be ignored. 
     */  
    virtual status_t        takePicture(int msgType) = 0;  
  
    // set preview/capture parameters - key/value pairs  
    virtual status_t        setParameters(const String8& params) = 0;  
   
    // get preview/capture parameters - key/value pairs  
    virtual String8         getParameters() const = 0;  
  
    // send command to camera driver  
    virtual status_t        sendCommand(int32_t cmd, int32_t arg1, int32_t arg2) = 0;    
    // tell the camera hal to store meta data or real YUV data in video buffers.  
    virtual status_t        storeMetaDataInBuffers(bool enabled) = 0;  
};

ICamera接口有两个子类BnCamera和BpCamera，是Binder通信的两端，BpCamera提供客户端调用 接口，BnCamera封装具体的实现，BnCamera也并没有真正实现ICamera相关接口而是在BnCamera子类 CameraService::Client中进行实现。而在CameraService::Client类中会继续调用硬件抽象层中相关方法来具体实现 Camera功能