asked on

Hbitmap to RGB Buffer help

Hello,

inside an external DLL I have a function to set a RGB Buffer for a logo to render. This function is:

 void       SetLogo(unsigned char *RGBBuf, int width, int strideLen, int height); // sets the actual logo

Open in new window

The description of this function is:
RGBBuf is a bitmap buffer, which contains a large bitmap width * height. Because Windows bitmaps still have padding inside, one must also specify strideLen in bytes, that is the line size. Without padding the line size is width * 3, with padding it varies between width * 3 + 0 and width * 3 + 3

I use now following code:

void CDVDConverterWork::GetImageForLogo()
{
	HBITMAP hBitmap = (HBITMAP)LoadImage(GetModuleHandle(NULL), MAKEINTRESOURCE(IDB_OVLLOGO), IMAGE_BITMAP, 0, 0, 0); 

	BITMAP bm;
	::GetObject( hBitmap , sizeof(bm) , &bm );

	/* Omitting error checks for brevity */
	HDC dcBitmap = CreateCompatibleDC ( NULL );
	SelectObject( dcBitmap, hBitmap );

	BITMAPINFO bmpInfo;
	bmpInfo.bmiHeader.biSize = sizeof(BITMAPINFOHEADER);
	bmpInfo.bmiHeader.biWidth = bm.bmWidth;
	bmpInfo.bmiHeader.biHeight = -bm.bmHeight;
	bmpInfo.bmiHeader.biPlanes = 1;
	bmpInfo.bmiHeader.biBitCount = 24;
	bmpInfo.bmiHeader.biCompression = BI_RGB;        
	bmpInfo.bmiHeader.biSizeImage = 0;  

	 
	unsigned char* retArray = new unsigned char[bmpInfo.bmiHeader.biSizeImage];

	//LPVOID retArray = malloc(bmpInfo.bmiHeader.biSizeImage);
	GetDIBits( dcBitmap , hBitmap , 0 , bm.bmHeight , retArray , &bmpInfo , DIB_RGB_COLORS );

	
	m_pConverter->SetLogo(retArray,bm.bmWidth, bmpInfo.bmiHeader.biSizeImage, bm.bmHeight);
	m_pConverter->EnableLogo(true);
	
}

Open in new window

But I get now crashes inside the DLL. It seems that my function submit wrong data to the DLL. Any idea what is wrong there?

jkr

>>Any idea what is wrong there?

You're allocating an array of size 0:

bmpInfo.bmiHeader.biSizeImage = 0;  

unsigned char* retArray = new unsigned char[bmpInfo.bmiHeader.biSizeImage];

Open in new window

That can't work.

Korbus

Agree with jkr- perhaps you meant to use: bmpInfo.bmiHeader.biSize in your new statement?

Ingo Foerster

ASKER

Hi,,
ok, I changed the code:

	HBITMAP hBitmap = (HBITMAP)LoadImage(GetModuleHandle(NULL), MAKEINTRESOURCE(IDB_OVLLOGO), IMAGE_BITMAP, 0, 0, LR_CREATEDIBSECTION | LR_SHARED); 

	BITMAP bm;
	::GetObject( hBitmap , sizeof(bm) , &bm );

	CDC mem_dc;  
	mem_dc.CreateCompatibleDC(NULL);

	/* Omitting error checks for brevity */
	BITMAPINFO bmpInfo;
	bmpInfo.bmiHeader.biSize = sizeof(BITMAPINFOHEADER);
	bmpInfo.bmiHeader.biWidth = bm.bmWidth;
	bmpInfo.bmiHeader.biHeight = bm.bmHeight;
	bmpInfo.bmiHeader.biPlanes = bm.bmPlanes;
	bmpInfo.bmiHeader.biBitCount = bm.bmBitsPixel;
	bmpInfo.bmiHeader.biCompression = BI_RGB;        
	bmpInfo.bmiHeader.biSizeImage = 0;  
	bmpInfo.bmiHeader.biClrImportant = 0;
	bmpInfo.bmiHeader.biClrUsed = 0;

	unsigned char *Pixels = (unsigned char *)malloc(bm.bmWidth * bm.bmHeight * 4); //sizeof(unsigned char)

	::GetDIBits(mem_dc.m_hDC, hBitmap, 0, bm.bmHeight, Pixels, &bmpInfo, DIB_RGB_COLORS);
	//GetDIBits( hDC , hBitmap , 0 , bm.bmHeight , Pixels , &bmpInfo , DIB_RGB_COLORS );

	::DeleteObject(hBitmap);
	
	SetLogo(Pixels, bm.bmWidth, bm.bmWidth*3, bm.bmHeight);

Open in new window

But now I have a gray image that is mirrored and is slanted. Any idea what is wrong with my buffer?

ASKER CERTIFIED SOLUTION

jkr

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

Korbus

Not sure, but I suspect the params to the LoadImage function:

try LR_VGACOLOR: "Uses true VGA colors."
rather than LR_CREATEDIBSECTION : "When the uType parameter specifies IMAGE_BITMAP, causes the function to return a DIB section bitmap rather than a compatible bitmap."

(from : http://msdn.microsoft.com/en-us/library/windows/desktop/ms648045(v=vs.85).aspx)

Ingo Foerster

ASKER

Was the passing issue. Thanks.

sarabande

unsigned char *Pixels = (unsigned char *)malloc(bm.bmWidth * bm.bmHeight * 4);

Open in new window

instead of multiplying times 4 you better should use (bm.bmBitsPixel/8) and instead of a separate pointer Pixels you better have used the bmiColors member array of structure BITMAPINFO.

in case of BI_RGB you have 32 bits per pixel what is 4 bytes what is the size of a RGBQUAD what is 3 bytes for red, green, blue + 1 byte reserved which could be used for specials like alpha blending. an RGBQUAD which has size of int on most systems is more handy than a RGBTRIPLE and the 4th byte could be well used for additional purposes.

nevertheless, when working with bitmaps you never should rely on assumptions but use the information given in the appropriate structures, here in the Bitmap structure.

The BITMAPINFO should have dynamic size such that the bmiColors array has appropriate size:

size_t sizePixels = 0;
BITMAPINFO * pBmpInfo = NULL;
...
sizePixels = bm.bmWidth * bm.bmHeight * (bm.bmBitsPixel/8);
// the structure was created with a size such that the 
// bmiColors member could take the required colors per pixel.
pBmpInfo = (unsigned char *)malloc( (sizeof(BITMAPINFOHEADER) + sizePixels);

pBmpInfo->bmiHeader.biWidth = bm.bmWidth;
pBmpInfo->bmiHeader.biHeight = bm.bmHeight;
...
::GetDIBits(mem_dc.m_hDC, hBitmap, 0, bm.bmHeight, pBmpInfo->bmiColors, &bmpInfo, DIB_RGB_COLORS);

Open in new window

btw, you seem to use the c++ compiler and not the c compiler. if so, you should use new/delete rather than malloc/free.

Sara