Dialog improvements

*1 Frame expandable:*

In this idea, we would like to change the frame dimensions automatically if the user enters an extra line of text. We call 'frame-expand-height' in Abiword. In fact, abiword already has this feature but does not allow users to define it or change it from within the abiword User Interface.
So my suggestions are:
1. expose 'frame-expand-height' to users when users use 'Frame' in abiword, since I think 'frame-expand-height' is more friendly to users.
2. in UI (Dialog : Format text box), users can turn it on or off
      on : enable frame-expand-height
      off: disable frame-expand-height

in code level:
in function: void FV_FrameEdit::mouseRelease(UT_sint32 x, UT_sint32 y)
we need return a valid value for "frame-expand-height",m_sExpandHeight.c_str()'

Steps to do this:
Firstly, all properities are defined in the array AbiPropertyName.
static const char * AbiPropertyName[PP_PropertyMap::abi__count] = {
       "frame-col-xpos",
       "frame-col-ypos",
       "frame-expand-height",
       "frame-height",
       "frame-horiz-align",
       "frame-min-height",
       "frame-page-xpos",
       "frame-page-ypos",
       "frame-pref-column",
       "frame-pref-page",
       "frame-rel-width",
       "frame-type",
       "frame-width",
}

and then in static PP_Property _props[], we init them to:

 * Property Nme: Initial Value: Can Inherit: Pointer to class : tPropLevel
 * tPropLevel should be set by or-ing the values defined in PP_Property.h
{"frame-expand-height",    "0.0in",           false, NULL, PP_LEVEL_FRAME},
{"frame-height",           "0.0in",           false, NULL, PP_LEVEL_FRAME},
{"frame-horiz-align",      "left",            false, NULL, PP_LEVEL_FRAME},
{"frame-min-height",       "0.0in",           false, NULL, PP_LEVEL_FRAME},
{"frame-page-xpos",        "0.0in",           false, NULL, PP_LEVEL_FRAME},

when user change the frame, presently they drag and release the frame,
the following action happens:
1.  ap_EditMethods::releaseFrame(AV_View * pAV_View, EV_EditMethodCallData * )
2.  void FV_View::releaseFrame
3.  bool pt_PieceTable::_fmtChangeStruxWithNotify
4.  fl_ContainerLayout::lookupProperties()
5.  in the lookupProperties function:
           1, we get Frame Type:  pSectionAP->getProperty("frame-type",pszFrameType))
           2, we get Position-to value:
pSectionAP->getProperty("position-to",pszPositionTo))
           3,  also: wrap-mode      wrap-mode
           4, ...

the one about this idea:
       if(pSectionAP &&
pSectionAP->getProperty("frame-expand-height",pszExpandHeight))
       {
               m_iMinHeight = m_iHeight;
               m_bExpandHeight = true;
       }

at this time, Abiword can receive "frame-expand-height" after user
enter extra lines. and the two values will be used to refresh Frame's
dimensions:
        m_iMinHeight = m_iHeight;
        m_bExpandHeight = true;

 2 Same as Frame expandable, all the attributions are stored in:

  static const char * AbiPropertyName[PP_PropertyMap::abi__count] = {

      "annotation-author",

      "annotation-date",

      "annotation-title",

      "background-color",

      "background-image",

      ………..

  };

Including "table-width", we can also need to add 'rotate-angle'.

 in AP_Dialog_FormatTable, we will correct all attribution values and store them in vector. (UT_PropVector m_vecProps;)

Example: setBorderColor : background-color:

        m_vecProps.addOrReplaceProp ("background-color", bgcol.c_str ());

 
Steps to implement:

So that in order to add the priority of setting the width and height. We need to add width and height to the vector.

1.  Add evoke function in AP_Dialog_FormatTable

2.  In the evoke function, setting corresponding value to the UT_PropVector.

3.  These two functions will render the attributions:

    bool pt_PieceTable::_realChangeStruxFmt

    bool pt_PieceTable::_fmtChangeFmtMark

 Add 'rotate-angle', we need to do:

1. Add attribution in static const char * AbiPropertyName[PP_PropertyMap::abi__count]

2. Init this attribution in static PP_Property _props[], we init them to:

   * Property Nme: Initial Value: Can Inherit: Pointer to class : tPropLevel

   * tPropLevel should be set by or-ing the values defined in PP_Property.h

3.  the most important step:

    Add render function to rotate-angle (this relate with another GSOC topic, I also have some interesting to implement it)

4.  Add evoke function in AP_Dialog_FormatTable

5.  In the evoke function, setting corresponding value to the UT_PropVector.

6.  These two functions will render the attributions:

     bool pt_PieceTable::_realChangeStruxFmt

• The project is a sub project of the national 973 project’ INTERACTVE TECHNOLOGY AND INTERFACE SUPPORT VIRTURAL REALITY MIXED ERVIRONMENT ’.Which is to use video instead of traditional keyboard/mouse interactive method.
• The technology involved in project include face tracking, combined with the convergence method of face skin color and characteristics, use AdaBoost face detection algorithm to detect face’s initial location, use Meanshift tracking algorithm to locate face’s position in each frame, calculate the direction of the movement and distance through the face offset displacement, send the command to the system instead of keyboard and mouse, this project can be applied to daily entertainment game.
• Development environment: C++, Open CV, Windows XP
• Responsibility 1: using detection and tracking at the same time to face tracking.
• 1.In order to accurately real-time tracking the target, using AdaBoost face detection algorithm  and CamShift tracking algorithm at the same time, start tracking algorithm to locate the position of the face and calculates the displacement when detection algorithm failed; start detection algorithm in the sector when the target tracking failed. 
• 2.Optimize AdaBoost face detection algorithm through introducing mechanism such as merging effective detecting windows and time impression window and so on, to get more accurate face location information and tracking scale information. 
• Responsibility 2:  revising Meanshift algorithm tracking window scale by Graph-cut theory.
• 1.The classical Mean Shift can’t change the scale of tracking window in real time while tracking target is changing in size. Graph cuts theory will separate out different characteristics of objects. according to this, graph cuts theory can be used to real-time separate out skin color piece to update tracking result scale. 
• 2.The paper "Scale of Shift Based on Adaptation since a Graph keep "is published in the international conference and indexed by EI.

• This project is a sub project of national project 863 ‘MULTIMODAL INTERFACE INPUT FLOW FUSION AND DISPLAY TECHNOLOGY IN VIRTUAL ENVIRONMENT’.
• It aims to get the relative displacements between the phone and the user, which are then mapped to the user’s operations to the large screen. So the remote large screen can be controlled by the user.
• It extracts relative displacement information from the video based on different visual methods (motion estimation, and face tracking), and amends or accelerates the existing calculation with sensor data. It improves user experience with the phone’s touch screen.
• Development environment: C++, Open CV, Windows XP
• Responsibility: Extracting the relative displacement of smart phone.
• 1.Input preprocessing: convert the image captured from camera tor YCbCr_420_SP format, JPG Code word throttling, get throttling after format conversion and compression coding. Transmit to the server and unzip a Bitmap image.
• 2.Estimating the displacement of the camera by using a similar tracking motion estimation method: converting the extracted  Bitmap image to gray image, introducing pyramid sampling mechanism, extracting the center of reference template as the core sub figure, with all the scope, computing the image displacement between the current frame and before the frame after sampling, and extending the scope of tracking extends to the image boundaries
• and smooth filtering the tracking result, thus get camera displacement estimation algorithm which is suitable for mobile phone big screen of interactive system.

• Job Description: our testing object is a lightweight browser based on the core chrome object on the phone, thin client mechanism can guarantee the mobile browser speed
• Job Responsibility: In the team, I am responsible for helping developers to unit test by using Google's open source c + + unit Test framework Google Test (GTest).According to different function to write the corresponding test cases and record failure log .

• Job Responsibility: I am responsible for the integration and maintenance of the department's service sites using IBM’s web development platform, and I update CDL service department web site and unit web page style.

• Proficient in C/C++, Experienced in Visual C++, MFC programming and the usage of OpenCV library.
• Familiar with data structure and algorithms, Experienced in Emgu、WPF.
• Experienced in programming in Linux and Experienced in gcc/gdb.
• Familiar with fundamental principles of OS. Familiar with Visual Studio、OpenCV