.NET (Core and Framework)

15-Jul-05 4:03

I'm not sure whether the TIFF files need to handled any differently. I believe you can directly instantiate an Image object using the FromFile() method and use it howsoever you desire.

Image Binarization

sarabjs13-Jul-05 7:07

13-Jul-05 7:07

I load an image, convert it to bitmapData using the 24bppRGB pixel format, binarize it using a simple routine (shown below) and save it using the Image.Save() method. The image saved however is still 24bppRGB - although all colours have the same value, either 0 or 255 - and shows up as an RGB image in Photoshop.

Any ideas on how I can change it to binary, ie a 1bpp format?

Code:
------

Bitmap bMap=new Bitmap (iMage);
BitmapData bData = bMap.LockBits(new Rectangle(0, 0, bMap.Width, bMap.Height), ImageLockMode.ReadWrite, PixelFormat.Format24bppRgb);
System.IntPtr Scan0 = bData.Scan0;
int iStride=bData.Stride;
int iht=bMap.Height;
int iwd=bMap.Width;
int iPix=new int[iht][];
unsafe
{
byte *p = (byte*)(void*)Scan0;
int nOffset = iStride-iwd*3;
for(int y=0;y<iht;++y)
{
iPix[y]=new int[iwd];
for(int x=0;x<iwd;++x)
{
if (conditionA) p[0]=p[1]=p[2]=0;
else p[0]=p[1]=p[2]=255;
p += 3;
}
p += nOffset;
}
}
bMap.UnlockBits(bData);

bMap.Save(fileName);

Re: Image Binarization

Christian Graus13-Jul-05 12:41

13-Jul-05 12:41

You need to convert it to a different format. It won't magically detect how many colours you've used and therefore change itself. There is a Clone method that takes a bit depth, but I think it's buggy. You can create a new image of the required bit depth and copy your old image onto the new. Or you could create a new image to start with of the right bit depth, and set it's values instead of setting values on your existing image.

Christian Graus - Microsoft MVP - C++

sarabjs14-Jul-05 3:55

14-Jul-05 3:55

Thanks Christian,

I found a solution from: http://www.bobpowell.net/onebit.htm

Here's what I did:

// new bitmap
Bitmap bm=new Bitmap(iwd, iht,PixelFormat.Format1bppIndexed);
BitmapData bmdn=bm.LockBits(new Rectangle(0,0,bm.Width,bm.Height),ImageLockMode.ReadWrite,PixelFormat.Format1bppIndexed);

// variables
int iIndex;
byte bP, bMask;

for(int y=0;y<iht;++y)
{
for(int x="0;x<iwd;++x)
{
iIndex=y*bmdn.Stride+(x">>3);
bP=Marshal.ReadByte(bmdn.Scan0,iIndex);
bMask=(byte)(0x80>>(x&0x7));
if (ConditionA) bP|=bMask;
else bP &=(byte)(bMask^0xff);
Marshal.WriteByte(bmdn.Scan0,iIndex,bP);
}
}
bm.UnlockBits(bmdn);

Christian Graus14-Jul-05 11:11

14-Jul-05 11:11

What is ConditionA ? This looks to me like it could be a lot slower than your old code ?

Either way, it's what I said to do - create a 1 bpp image to write to.

Christian Graus - Microsoft MVP - C++

sarabjs15-Jul-05 3:58

15-Jul-05 3:58

I have a 2-dimensional integer array of the same width and height of the image, each of whose elements is either 1 or 0. Condition A just checks whether the corresponding element in the array is a 1 or a 0, and sets the pixel accordingly

There doesn't seem to be a major drop in the time taken by the overall routine though. Do you have any suggestions?

Sarabjit.

Christian Graus17-Jul-05 12:58

17-Jul-05 12:58

If it's just an array lookup, then it probably is fine, I thought it may be missing () and be a function call that did goodness knows what.

Christian Graus - Microsoft MVP - C++

sarabjs20-Jul-05 5:07

How to change the SourcePath in a Microsoft Installer project?

20-Jul-05 5:07

Yea, it is just an array lookup...

Copying Object Types

SamM13-Jul-05 6:14

SamM

13-Jul-05 6:14

Hi
I am working on a project where I have a class containing a field of type object. I need to get a deep copy of the class, but cannot see an obvious way to copy the value of the field whos type is declared as object. I have considered using reflection to try and achieve this, but was wondering if anyone knew of a simpler way.

Any suggestions would be greatly appreciated.

Regards
Sam

Re: Copying Object Types

S. Senthil Kumar13-Jul-05 6:43

S. Senthil Kumar

13-Jul-05 6:43

Just serialize and deserialize it out of a MemoryStream. The framework takes care of the rest. Something like

using (MemoryStream s = new MemoryStream())
{
	BinaryFormatter f = new BinaryFormatter();
	f.Serialize(s, obj);
	
	s.Seek(0, SeekOrigin.Begin);
	object o = new BinaryFormatter().Deserialize(s);
	Console.WriteLine(Object.ReferenceEquals(o, obj));
}

Regards
Senthil
_____________________________
My Blog | My Articles | WinMacro

Re: Copying Object Types

SamM13-Jul-05 22:30

SamM

13-Jul-05 22:30

Thanks Senthil, great help.

Regards
Sam

bondguru12-Jul-05 20:18

bondguru

12-Jul-05 20:18

Hi all,
I am working on Microsoft Installer project to create a setup for a software. The project is already existing and working. However now I have changed the folder hierarchy i.e. I have now changed the folders from where the files to be packaged are extracted. In the project, in Properties page there is an entry SourcePath which displays the path where the file is present. This entry is still showing the old path of the files. I can't change it's contents as it is read-only. So is there any other way to change the path? If so please inform me.
Thanking You.

Regards,
Harshad

Re: How to change the SourcePath in a Microsoft Installer project?

rwestgraham13-Jul-05 15:22

rwestgraham

13-Jul-05 15:22

No. The VS.NET Setup creation IDE is very limited and poorly designed. The only way to change the source files is to delete the existing ones from the setup project and add the new ones.

Robert

Mysterious application crash

RajeshGuptha12-Jul-05 5:33

RajeshGuptha

12-Jul-05 5:33

Dear all,

My .Net WinForms Application quietly closes off without a warning and I am unable to understand the root cause of the problem, as it happens sporadically.

I am working with .Net Framework 1.1 on Windows XP (tested on both SP 1 and 2).

I tried the following:
- Having a try/catch cover on all the important event handlers (OnPaint, OnMouse..., etc) and in the static Main() method (Application.Run(...))
- Set an UnhandledException handler for the present App domain.
- Earlier, I was creating and releasing a COM object in a method that was being called many times. Now, I moved the creation into the class constructor and the releasing into the Dispose().

Yet, none of these options have either stopped the crash or have given me a clue.

Is there a way in which I can get hold of any low level API's or tools that will be able to execute my application in a shell, which will give me a greater control on locating the cause of the crash.

Any ideas will be greatly appreciated.

Thanks for reading.

-Rajesh.

Re: Mysterious application crash

Christian Graus12-Jul-05 13:34