set( ANDROID_ABI "${ANDROID_ABI}" CACHE STRING "The target ABI for Android. If arm, then armeabi-v7a is recommended for hardware floating point." FORCE )
@@ -75,7 +75,7 @@ You need the following software to be installed in order to develop for Android
sudo update-java-alternatives --set java-6-sun
**TODO:** add a note on Sun/Oracle Java installation on Ubuntu 12.
.. **TODO:** add a note on Sun/Oracle Java installation on Ubuntu 12.
#. **Android SDK**
...
...
@@ -241,6 +241,7 @@ where:
The script :file:`Android.mk` usually has the following structure:
.. code-block:: make
:linenos:
LOCAL_PATH := $(call my-dir)
...
...
@@ -258,6 +259,7 @@ This is the minimal file :file:`Android.mk`, which builds C++ source code of an
Usually the file :file:`Application.mk` is optional, but in case of project using OpenCV, when STL and exceptions are used in C++, it also should be created. Example of the file :file:`Application.mk`:
.. code-block:: make
:linenos:
APP_STL := gnustl_static
APP_CPPFLAGS := -frtti -fexceptions
...
...
@@ -337,7 +339,7 @@ We recommend the approach based on Eclipse :abbr:`CDT(C/C++ Development Tooling)
:alt: Configure CDT
:align: center
` `
And:
.. image:: images/eclipse_cdt_cfg2.png
:alt: Configure CDT
...
...
@@ -350,6 +352,7 @@ We recommend the approach based on Eclipse :abbr:`CDT(C/C++ Development Tooling)
:align: center
#. Open :guilabel:`Project Properties -> C/C++ Build`, unckeck ``Use default build command``, replace "Build command" text from ``"make"`` to
``"${NDKROOT}/ndk-build.cmd"`` on Windows,
``"${NDKROOT}/ndk-build"`` on Linux and MacOS.
...
...
@@ -393,7 +396,7 @@ We recommend the approach based on Eclipse :abbr:`CDT(C/C++ Development Tooling)
:alt: Configure CDT
:align: center
.. note:: The latest Android NDK **r8b** has a bit different STL headers path. So if you use this NDK version please use the following modified **Include** paths list:
.. note:: The latest Android NDK **r8b** uses different STL headers path. So if you use this NDK release add the following **Include** paths list instead:
::
...
...
@@ -412,12 +415,16 @@ AVD
AVD (*Android Virtual Device*) is not probably the most convenient way to test an OpenCV-dependent application, but sure the most uncomplicated one to configure.
#. Assuming you already have *Android SDK* and *Eclipse IDE* installed, in Eclipse go :guilabel:`Window -> AVD Manager`.
**TBD:** how to start AVD Manager without Eclipse...
.. **TBD:** how to start AVD Manager without Eclipse...
#. Press the :guilabel:`New` button in :guilabel:`AVD Manager` window.
#. :guilabel:`Create new Android Virtual Device` window will let you select some properties for your new device, like target API level, size of SD-card and other.
.. image:: images/AVD_create.png
:alt: Configure builders
:align: center
#. When you click the :guilabel:`Create AVD` button, your new AVD will be availible in :guilabel:`AVD Manager`.
#. Press :guilabel:`Start` to launch the device. Be aware that any AVD (a.k.a. Emulator) is usually much slower than a hardware Android device, so it may take up to several minutes to start.
#. Go :guilabel:`Run -> Run/Debug` in Eclipse IDE to run your application in regular or debugging mode. :guilabel:`Device Chooser` will let you choose among the running devices or to start a new one.
...
...
@@ -435,22 +442,31 @@ Windows host computer
#. Attach the Android device to your PC with a USB cable.
#. Go to :guilabel:`Start Menu` and **right-click** on :guilabel:`Computer`. Select :guilabel:`Manage` in the context menu. You may be asked for Administrative permissions.
#. Select :guilabel:`Device Manager` in the left pane and find an unknown device in the list. You may try unplugging it and then plugging back in order to check whether it's your exact equipment appears in the list.
.. image:: images/usb_device_connect_01.png
:alt: Unknown device
:align: center
#. Try your luck installing `Google USB drivers` without any modifications: **right-click** on the unknown device, select :guilabel:`Properties` menu item --> :guilabel:`Details` tab --> :guilabel:`Update Driver` button.
.. image:: images/usb_device_connect_05.png
:alt: Device properties
:align: center
#. Select :guilabel:`Browse computer for driver software`.
.. image:: images/usb_device_connect_06.png
:alt: Browse for driver
:align: center
#. Specify the path to :file:`<Android SDK folder>/extras/google/usb_driver/` folder.
.. image:: images/usb_device_connect_07.png
:alt: Browse for driver
:align: center
#. If you get the prompt to install unverified drivers and report about success - you've finished with USB driver installation.
.. image:: images/usb_device_connect_08.png
:alt: Install prompt
:align: center
...
...
@@ -460,23 +476,33 @@ Windows host computer
.. image:: images/usb_device_connect_09.png
:alt: Installed OK
:align: center
#. Otherwise (getting the failure like shown below) follow the next steps.
.. image:: images/usb_device_connect_12.png
:alt: No driver
:align: center
#. Again **right-click** on the unknown device, select :guilabel:`Properties --> Details --> Hardware Ids` and copy the line like ``USB\VID_XXXX&PID_XXXX&MI_XX``.
.. image:: images/usb_device_connect_02.png
:alt: Device properties details
:align: center
#. Now open file :file:`<Android SDK folder>/extras/google/usb_driver/android_winusb.inf`. Select either ``Google.NTx86`` or ``Google.NTamd64`` section depending on your host system architecture.
.. image:: images/usb_device_connect_03.png
:alt: "android_winusb.inf"
:align: center
#. There should be a record like existing ones for your device and you need to add one manually.
.. image:: images/usb_device_connect_04.png
:alt: "android_winusb.inf"
:align: center
#. Save the :file:`android_winusb.inf` file and try to install the USB driver again.
.. image:: images/usb_device_connect_05.png
:alt: Device properties
:align: center
...
...
@@ -492,7 +518,9 @@ Windows host computer
.. image:: images/usb_device_connect_07.png
:alt: Browse for driver
:align: center
#. This time installation should go successfully.
.. image:: images/usb_device_connect_08.png
:alt: Install prompt
:align: center
...
...
@@ -502,11 +530,15 @@ Windows host computer
.. image:: images/usb_device_connect_09.png
:alt: Installed OK
:align: center
#. And an unknown device is now recognized as an Android phone.
.. image:: images/usb_device_connect_10.png
:alt: "Known" device
:align: center
#. Successful device USB connection can be verified in console via ``adb devices`` command.
.. image:: images/usb_device_connect_11.png
:alt: "adb devices"
:align: center
...
...
@@ -523,7 +555,7 @@ By default Linux doesn't recognize Android devices, but it's easy to fix this is
Then restart your adb server (even better to restart the system), plug in your Android device and execute :command:`adb devices` command. You will see the list of attached devices:
@@ -57,7 +57,7 @@ Using async initialization is a **recommended** way for application development.
To run OpenCV Manager-based application the first time you need to install packages with the `OpenCV Manager` and `OpenCV binary pack` for you platform.
You can do it using Google Play Market or manually with ``adb`` tool:
@@ -327,11 +331,10 @@ It will be capable of accessing camera output, processing it and displaying the
// TODO setup camera
}
..
#. Create a new *Activity* (*New -> Other -> Android -> Android Activity*) and name it, for example: *HelloOpenCVActivity*. For this activity define *onCreate*, *onResume* and *onPause* voids.
.. code-block:: java
:linenos:
public void onCreate (Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
...
...
@@ -359,11 +362,12 @@ It will be capable of accessing camera output, processing it and displaying the
});
ad.show();
}
}
#. Add the following permissions to the AndroidManifest.xml file:
.. code-block:: xml
:linenos:
</application>
...
...
@@ -372,12 +376,15 @@ It will be capable of accessing camera output, processing it and displaying the
Class used for background/foreground segmentation. ::
Class used for background/foreground segmentation. ::
class GMG_GPU_GPU
{
...
...
@@ -677,9 +677,9 @@ Class used for background/foreground segmentation. ::
...
};
The class discriminates between foreground and background pixels by building and maintaining a model of the background. Any pixel which does not fit this model is then deemed to be foreground. The class implements algorithm described in [GMG2012]_.
The class discriminates between foreground and background pixels by building and maintaining a model of the background. Any pixel which does not fit this model is then deemed to be foreground. The class implements algorithm described in [GMG2012]_.
Here are important members of the class that control the algorithm, which you can set after constructing the class instance:
Here are important members of the class that control the algorithm, which you can set after constructing the class instance:
uchar4 dst_data = convert_uchar4_sat_rte((convert_float4(a))* ud * vd +(convert_float4(b))* u * vd + (convert_float4(c))* ud * v + (convert_float4(d)) * u * v );
