how to match two image signatures

The details depend on the background of the signature images and the location of the signature in each image.
For example, if you are dealing with structured documents that has a specific location for signatures, the background color of the signature is known, the width and height of the image is fixed, you can do this as follows:
- Auto crop each image to remove any extra space around the signature.
- Deskew each image to straighten the contents of the image.
- Combine the two images with XOR factor. The XOR factor means the different pixels gives white color.
- Count the white pixels in the original and the result (combined image).
- If the difference (in the number of white pixels) between the original and the combined image is small, you can consider the two images are similar. Otherwise. If the difference is large, this means the two images are different.

LEADTOOLS Programming toolkit supports almost all the above mentioned image processing features. You can find more details with sample codes here:
https://www.leadtools.com/help/leadtools/v19/dh/po/autocropcommand.html
https://www.leadtools.com/help/leadtools/v19/dh/po/deskewcommand.html
https://www.leadtools.com/help/leadtools/v19/dh/pe/combinecommand.html
https://www.leadtools.com/help/leadtools/v19/dh/pc/colorcountcommand.html

The following code suggests how you can do this using LEADTOOLS SDK:

using (RasterCodecs codecs = new RasterCodecs())
{
_ocrimage = codecs.Load(filename);
RasterImage _source1 = codecs.Load(@"c:\image1.png");
RasterImage _source2 = codecs.Load(@"c:\images.png");
//Prepare the command
//AutoCrop the image with 20 tolerance
Leadtools.ImageProcessing.Core.AutoCropCommand _autoCropCommand = new Leadtools.ImageProcessing.Core.AutoCropCommand();
_autoCropCommand.Threshold = 20;
_autoCropCommand.Run(_source1);
_autoCropCommand.Run(_source2);
//Prepare the command
//Deskew the image
Leadtools.ImageProcessing.Core.DeskewCommand _deskewCommand = new Leadtools.ImageProcessing.Core.DeskewCommand();
_deskewCommand.Flags = Leadtools.ImageProcessing.Core.DeskewCommandFlags.DeskewImage | Leadtools.ImageProcessing.Core.DeskewCommandFlags.DoNotFillExposedArea;
_deskewCommand.Run(_source1);
Leadtools.ImageProcessing.Effects.CombineCommand _combineCommand = new Leadtools.ImageProcessing.Effects.CombineCommand();
_combineCommand.DestinationRectangle = new LeadRect(0, 0, _source1.Width, _source1.Height);
_combineCommand.SourcePoint = new LeadPoint(0, 0);
_combineCommand.SourceImage = _source1;
_combineCommand.Flags = Leadtools.ImageProcessing.Effects.CombineCommandFlags.OperationXor;
_combineCommand.Run(_source2);
//Add a region to each image so that it contains only white pixels
_source1.AddColorToRegion(RasterColor.White, RasterRegionCombineMode.Set);
_source2.AddColorToRegion(RasterColor.White, RasterRegionCombineMode.Set);
int colorCountSource = 0;
int colorCountResult = 0;
//Prepare the command
//Get the number of colors in the image
Leadtools.ImageProcessing.Color.ColorCountCommand _colorCountCommand = new Leadtools.ImageProcessing.Color.ColorCountCommand();
_colorCountCommand.Run(_source1);
colorCountSource = _colorCountCommand.ColorCount;
_colorCountCommand.Run(_source2);
colorCountResult = _colorCountCommand.ColorCount;
bool IsSimilar = false;
if (colorCountResult - colorCountSource < 100)
IsSimilar = true;
if (IsSimilar == true)
Console.WriteLine("The two images are similar");
else
Console.WriteLine("The two images are different");
}

2 Replies

Know the answer? Post it — somebody with the same question will find it here.

Maen Badwan8y ago
The details depend on the background of the signature images and the location of the signature in each image.
For example, if you are dealing with structured documents that has a specific location for signatures, the background color of the signature is known, the width and height of the image is fixed, you can do this as follows:
- Auto crop each image to remove any extra space around the signature.
- Deskew each image to straighten the contents of the image.
- Combine the two images with XOR factor. The XOR factor means the different pixels gives white color.
- Count the white pixels in the original and the result (combined image).
- If the difference (in the number of white pixels) between the original and the combined image is small, you can consider the two images are similar. Otherwise. If the difference is large, this means the two images are different.

LEADTOOLS Programming toolkit supports almost all the above mentioned image processing features. You can find more details with sample codes here:
https://www.leadtools.com/help/leadtools/v19/dh/po/autocropcommand.html
https://www.leadtools.com/help/leadtools/v19/dh/po/deskewcommand.html
https://www.leadtools.com/help/leadtools/v19/dh/pe/combinecommand.html
https://www.leadtools.com/help/leadtools/v19/dh/pc/colorcountcommand.html

The following code suggests how you can do this using LEADTOOLS SDK:
using (RasterCodecs codecs = new RasterCodecs())
{
   _ocrimage = codecs.Load(filename);

   RasterImage _source1 = codecs.Load(@"c:\image1.png");
   RasterImage _source2 = codecs.Load(@"c:\images.png");

   //Prepare the command
   //AutoCrop the image with 20 tolerance
   Leadtools.ImageProcessing.Core.AutoCropCommand _autoCropCommand = new Leadtools.ImageProcessing.Core.AutoCropCommand();
   _autoCropCommand.Threshold = 20;
   _autoCropCommand.Run(_source1);
   _autoCropCommand.Run(_source2);

   //Prepare the command
   //Deskew the image
   Leadtools.ImageProcessing.Core.DeskewCommand _deskewCommand = new Leadtools.ImageProcessing.Core.DeskewCommand();
   _deskewCommand.Flags = Leadtools.ImageProcessing.Core.DeskewCommandFlags.DeskewImage | Leadtools.ImageProcessing.Core.DeskewCommandFlags.DoNotFillExposedArea;
   _deskewCommand.Run(_source1);

   Leadtools.ImageProcessing.Effects.CombineCommand _combineCommand = new Leadtools.ImageProcessing.Effects.CombineCommand();
   _combineCommand.DestinationRectangle = new LeadRect(0, 0, _source1.Width, _source1.Height);
   _combineCommand.SourcePoint = new LeadPoint(0, 0);
   _combineCommand.SourceImage = _source1;
   _combineCommand.Flags = Leadtools.ImageProcessing.Effects.CombineCommandFlags.OperationXor;
   _combineCommand.Run(_source2);

   //Add a region to each image so that it contains only white pixels
   _source1.AddColorToRegion(RasterColor.White, RasterRegionCombineMode.Set);
   _source2.AddColorToRegion(RasterColor.White, RasterRegionCombineMode.Set);

   int colorCountSource = 0;
   int colorCountResult = 0;

   //Prepare the command
   //Get the number of colors in the image
   Leadtools.ImageProcessing.Color.ColorCountCommand _colorCountCommand = new Leadtools.ImageProcessing.Color.ColorCountCommand();
   _colorCountCommand.Run(_source1);
   colorCountSource = _colorCountCommand.ColorCount;

   _colorCountCommand.Run(_source2);
   colorCountResult = _colorCountCommand.ColorCount;

   bool IsSimilar = false;

   if (colorCountResult - colorCountSource < 100)
      IsSimilar = true;

   if (IsSimilar == true)
      Console.WriteLine("The two images are similar");
   else
      Console.WriteLine("The two images are different");
}
0
Nilesh Patil9y ago
Hi,

Refer this links

http://www.c-sharpcorner.com/uploadfile/prathore/image-comparison-using-C-Sharp/

https://blogs.msdn.microsoft.com/domgreen/2009/09/06/comparing-two-images-in-c/
-1