Jump to content
SimTheNo1

_GDIPlus_ImageLoadFromFile or _GDIPlus_BitmapCreateFromFile ?

Recommended Posts

SimTheNo1

I am busy with building a solution for change monitoring of VOIP call program and to be properly automated means among others need for some limited OCR functionality. Current works of others are way too much overkill for this case what makes the need to build it myself. But to do it properly I significantly have to increase my knowledge about digital graphics management.

For now I already have discovered here and there some mind blowing  Autoit miracles what can be achieved with Windows own possibilities to manipulate that what is output to the monitor. And as far I can judge there are 2 options to process graphics without use of any external libraries like ImageMagick, FreeImage and so on. These are:

  1.  WinAPI
  2.  GDIPlus

It is for me quite obvious to have various holes in general understanding of graphics and it is once more very clear what advantages gives consistent general study in 1 or another official institutions like University.  Cause there you are introduced into certain domain of knowledge in a way which usually has been perfected over long period of time. So you are not overloaded by stuff which requires a certain amount of information to be initially clear for you. For example, before starting to solve physics you first learn to read, count and so on and then move to subjects like physics.

Though in my case opportunity to study in such educational system I had only for 8 years, from my 7th to 15th year of age in the country that was falling apart now Ukraine but used to be USSR, was all what it was. After have immigrated to Netherlands possibilities to study further haven't occur. And this therefor causes often various implications when going deep in that or another field of practical knowledge acquiring for any needed physical result, like programming to perform enormous amount of tasks. In this particular case automating VOIP call program.

Anyway, right now I think the best direction to move is to concentrate on as basic as possible image management and if someone would maybe explain in general what is a pixel will definitely help. Particularly I am very curious about how to do picture manipulations in Autoit. Especially would help a lot to produce eventually following functions:

  • createImage($imageFileName, $width, $height, $color)
  • readImagePixel($imageFileName, $x, $y) 
  • writeImagePixel($imageFileName, $x, $y)

I do not know exactly how image is handled in computer but preferably above mentioned functions should deal with so far possible origin of graphics creation on computer. But nevertheless I definitely would love to hear any proposition for solution.

The problem with explaining screenshots:

  1.  VOIPConnect full Window VOIPConnect.jpg
  2. Part of Window with control to monitor for changes ActiveCallListCtrlRect.jpg
  3. Exact location of area where actual changes occur and have to be processed Area2Monitor.jpg

It comes down to a rectangle of approximately 51 pixel wide and 7 pixel high. In fact if I get to learn as far as to be able exactly read, write and compare 2 images consistently across different computers I could narrow down then even further the area to watch as little as a square of 2-5 pixels wide. 

To finish here is last detail about particularly no need for ultra fast solution at all. This because it is needed only once when it is first run on a new computer and to have to wait few minutes while it is being set instead of just a few seconds make no sence.

This is it and what I too think to do beside this very particular case is to purify out beautiful generic Autoit functions for core image manipulation by using WinAPI or/and GDIPlus. 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Similar Content

    • Xandy
      By Xandy
      MapIt is a tile world editor.  MapIt was built around the concept of reversing Dragon Warrior map images.  MapIt can take image input and produce a tile and world array.
      MapIt features Unity style dragable labels that adjust property values.

      MapParser is a C++ project that scans images for unique tiles.  MapParser is very fast.  Due to hard drive failure, many bugs were restored b/c I had to rewind many years.  Frustrated with the design, I wrote a new version from the ground up.  This New Version:  AutoIt Front-end, command line controls, and shared with the world; so that I can't lose it again.
      You can toggle the C++ MapParser off to see the difference in speeds between the MapParser CPP verse AutoIt function.  Function is named Scan_Tiles() in AutoIt.  You can also chose to download without MapParser.exe.
      At the moment Scanning a image resets the arrays, but you can add tiles after scanning.
      Images can be added as tiles without scan image at all.  Then configure settings to give your world parameters and manually fill the world data with tile indexes.
      Using the settings you can change tile size after a scan.  Example: you wanted to replace a map with different sized tiles.
      Changing and replacing tile / world data is easy.  B/c tile world editor.

      Hotkeys, I use CTRL+R in image above to signal replace tile action and I use "G" to Get the tile under mouse.
      Hotkeys are not saved to disk and thus are set to default between sessions.
      I might draw the world to pre-rendered surfaces and use them as multi-layer someday.  I do that in my AutoIt, DragonWarrior Remake but I could spend forever unsure what features are important for this.  The DW_Remake has a method of replacing a tile with a tile on two layers.  So you could replace a tree on the first layer with a grass, and a tree in the second layer.  This is all getting very confusing.
      I attempted to write the good code.  If something could be better, please advise.
      Fifth release.  Enjoy.
      For download, videos, and example of created world file data; please visit the MapIt webpage: http://songersoft.com/programming/mapit/mapit_about.phtml
      Special thanks: @AdmiralAlkex, @Melba23, @MrCreatoR
      Main AutoIt source file: Will not run without other Includes and SDL DLLs.
      Last Update: 5/26/2018 3:45 PM EST
      REMOVED CODE BLOCK:  I was informed the this page loaded very slowly, one solution so far has been to remove the 2k lines in the code block.
      When I recieve more feedback from the User I may reduce image size or remove images.
      Next Version Added: $eSETTING_TILE_LAST_PATH I have the weekend, I want to write world layers with aBoard surfaces.  
    • c.haslam
      By c.haslam
      I had thought that _GDIPlus_ImageClone($hImage) removes all property items, but I now know that it copies property items.
      What is the easiest way of copying an image that has property items to another image that does not have property items?
    • c.haslam
      By c.haslam
      I think that _GDIPlus_ImageGetPropertyItem should return a Property Item that can readily be set as a property of another image. Why would one want to do this? A script might (as one of mine does) edit an image of a .jpg file then write the result to a .jpg file. In editing it, in my case, the number of horizontal and vertical pixels change, and the date/time edited should be set as a property. All other property items remain the same. Other GDI+ users may change other properties.
      So an approach is to copy all the property items from image1 to image2, then update a few properties.
      M$ provides a Property Item class (id, length, pointer to an array of values) but _GDIPlus_ImageGetPropertyItem returns something different.
      I think that, wherever reasonable, UDFs that are wrappers for M$ methods should work like M$'s methods. _GDIPlus_ImageGetPropertyItem does not do this:
      While M$ returns 2 values for a value that is a ratio of 2 numbers,  _GDIPlus_ImageGetPropertyItem returns numerator/denominator as a single value (often a Double). _GDIPlus_ImageSetPropertyItem (when included in GDIPlus.au3) will be unable to set ratio property items properly because it cannot know what the numerator and denominator are. So copying such property items will not work properly. M$ has a Void* buffer where the buffer is an array of values but _GDIPlus_ImageGetPropertyItem() returns the values in the same 1-d array as id, length and type. M$'s approach produces a 1-d array with the same number of elements for all property items while _GDIPlus_ImageGetPropertyItem produces a 1-d array with various numbers of elements. To me, this is not good programming practice. For a photo from a Sony camera, the worst case has 1-d array with 67 elements. When combined into a 2-d array with the property items of all properties, there are 66 rows and 67 columns, with many elements not used. So I suggest that _GDIPlus_ImageGetPropertyItem look like this:
      ; #FUNCTION# ==================================================================================================================== ; Name ..........: cGDIPlus_ImageGetPropertyItem ; Description ...: Gets a specified property item (piece of meta data) from an Image object ; Syntax ........: cGDIPlus_ImageGetPropertyItem($hImage, $iPropID) ; Parameters ....: $hImage - Pointer to an image object ; $iPropID - Identifier of the property item to be retrieved ; Return values .: Success: Array containing the values of the property item ; [0] - identifier ; [1] - size, in bytes, of the value array ; [2] - type of value(s) in the value array ; [3] - value array ; Failure: Sets the @error flag to non-zero, @extended may contain GPSTATUS error code ($GPID_ERR*). ; Author ........: Eukalyptus ; Modified ......: c.haslam ; Remarks .......: types: unsigned byte = 1, ASCII string = 2, unsigned short = 3, unsigned long = 4, ; unsinged rational = 5, undefined = 7, signed long = 9, signed rational = 10 ; Related .......: _GDIPlus_ImageGetPropertyIdList ; Link ..........: https://msdn.microsoft.com/en-us/library/windows/desktop/ms535390(v=vs.85).aspx, ; https://msdn.microsoft.com/en-us/library/windows/desktop/ms534493(v=vs.85).aspx, ; https://msdn.microsoft.com/en-us/library/windows/desktop/ms534414(v=vs.85).aspx ; Example .......: Yes ; =============================================================================================================================== Func cGDIPlus_ImageGetPropertyItem($hImage, $iPropID) Local $iSize = __GDIPlus_ImageGetPropertyItemSize($hImage, $iPropID) If @error Then Return SetError(@error, @extended, -1) Local $tBuffer = DllStructCreate("byte[" & $iSize & "];") Local $pBuffer = DllStructGetPtr($tBuffer) Local $aResult = DllCall($__g_hGDIPDll, "int", "GdipGetPropertyItem", "handle", $hImage, "uint", _ $iPropID, "uint", $iSize, "struct*", $tBuffer) If @error Then Return SetError(@error, @extended, -1) If $aResult[0] Then Return SetError(10, $aResult[0], False) Local $tPropertyItem = DllStructCreate("int id; int length; short type; ptr pvalue;", $pBuffer) Local $iBytes = DllStructGetData($tPropertyItem, "length") Local $pValue = DllStructGetData($tPropertyItem, "pvalue") Local $aRet[4] Local $type = DllStructGetData($tPropertyItem,'type') Local $tValues, $iValues Switch $type Case 2 ;ASCII String $iValues = 1 $tValues = DllStructCreate("char[" & $iBytes & "];", $pValue) Case 3 ;Array of UShort $iValues = Int($iBytes / 2) $tValues = DllStructCreate("ushort[" & $iValues & "];", $pValue) Case 4, 5 ;Array of UInt / Fraction $iValues = Int($iBytes / 4) $tValues = DllStructCreate("uint[" & $iValues & "];", $pValue) Case 9, 10 ;Array of Int / Fraction $iValues = Int($iBytes / 4) $tValues = DllStructCreate("int[" & $iValues & "];", $pValue) Case Else ;Array of Bytes $iValues = 1 $tValues = DllStructCreate("byte[" & $iBytes & "];", $pValue) EndSwitch $aRet[0] = DllStructGetData($tPropertyItem,'id') $aRet[1] = $iBytes $aRet[2] = $type Local $aVals[$iValues] If $type=2 Or $type=7 Then ; ASCII string or undefined $aVals[0] = DllStructGetData($tValues,1) Else For $i = 0 To $iValues-1 $aVals[$i] = DllStructGetData($tValues,1,$i+1) Next EndIf $aRet[3] = $aVals Return $aRet EndFunc And here is an example:
      #include <GDIPlus.au3> #include <Array.au3> Example() Func Example() _GDIPlus_Startup() Local $hImage = _GDIPlus_ImageLoadFromFile(RegRead((@AutoItX64 = True ? "HKLM\SOFTWARE\Wow6432Node\AutoIt v3\AutoIt" : "HKLM\SOFTWARE\AutoIt v3\AutoIt"), "InstallDir") & "\Examples\GUI\Torus.png") If @error Then _GDIPlus_Shutdown() MsgBox(16, "", "An error has occured - unable to load image!", 30) Return False EndIf Local $ar = _GDIPlus_ImageGetPropertyIdList($hImage) Local $vPropNbrs[UBound($ar,1)-1] ; Extract ID numbers For $i = 1 To UBound($ar,1)-1 $vPropNbrs[$i-1] = $ar[$i][0] Next ; Get all property items Local $aPropItems[UBound($vPropNbrs)][4],$vPropItem For $i = 0 To UBound($vPropNbrs)-1 $vPropItem = cGDIPlus_ImageGetPropertyItem($hImage,$vPropNbrs[$i]) For $j = 0 To 3 $aPropItems[$i][$j] = $vPropItem[$j] Next Next ; Collapse values arrays so _ArrayDisplay can display them Local $ar = $aPropItems For $i = 0 To UBound($aPropItems,1)-1 $ar[$i][0] = '0x'&Hex($ar[$i][0],4) $ar[$i][3] = '' For $j = 0 To UBound($aPropItems[$i][3])-1 $ar[$i][3] &= ($aPropItems[$i][3])[$j]&'|' Next $ar[$i][3] = StringTrimRight($ar[$i][3],1) Next _ArrayDisplay($ar) _GDIPlus_Shutdown() EndFunc Unfortunately this example (based on one now in the Help) does not exercise most of the item types, but I do not know of a file with EXIF metadata that is on most PCs. I have tested this code by updating and adding property items to torus.jpg and to a photo taken by a Sony camera. The code for implementing and calling _GDIPlus_ImageSetPropertyItem will be fairly simple [see below].
       
      Your thoughts?
       
    • c.haslam
      By c.haslam
      I have a strange symptom of a problem somewhere in my code, or in a UDF: when I add code to get all properties of a GDI+ image to a 6000-line script,  the variable type of a parameter is reported as a pointer when the caller clearly has this argument as a string. If I comment out my code that gets the properties, the function properly sees the parameter as a string.
      [Edit: The reporting as a pointer and getting image properties are miles apart, both in where they are in the script code and in where they are run.]
      At this point, I am looking for clues as to why this is happening, so the hunt is on!
      I have noticed one odd thing, in _GDIPlus_ImageGetPropertyIdList() I see a line that begins Local $sPropertyTagInfo = . This "line" is split into 2 lines. SCiTE shows that the first line is 2454 characters and the second 2400. Adding these numbers I get 4854 characters.
      But the Help for AutoIt3 Limits/defaults says that MAXLINESIZE, Maximum size for a line of script, is 4096 characters.
      The line would exceed the limit if AutoIt considers the subject line to be 1 line and not 2 lines.
      My question: is this line legal? Is buffer overun possible?
    • c.haslam
      By c.haslam
      I find _GDIPlus_ImageClone in the Help, but neither in SciTe auto-complete nor in GDIPlus.au3
      Shoud this be reported in Trac?
      By searching the forum (and changing the name of the first arguement ot DLLCall, the code appears to be:
      Func _GDIPlus_ImageClone($hImage) Local $aResult = DllCall($__g_hGDIPDll, "uint", "GdipCloneImage", "handle", $hImage, "int*", 0) If @error Then Return SetError(@error, @extended, 0) Return $aResult[2] EndFunc ;==>_GDIPlus_ImageClone  
×

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.