/**
* Gets the traditional laplace graph of an image. The image
* is assumed to be able to be translated into black and white.
* The values in the graph can be negative.
*
* @param image The image to get the graph of.
* @return The graph of the image.
*/
public static double[][] GetGraph(Bitmap image)
{
var filter = new TraditionalLaplace(image, 0, 8);
//filter the image
var fparray =
new FilteredPixel[image.Width * image.Height];
for (var i = 0; i < image.Width; i++)
{
for (var j = 0; j < image.Height; j++)
{
fparray[i * image.Height + j] =
new FilteredPixel(i, j,
filter.GetValue(i, j));
}
}
//sort the filter results
//is in ascending order - low at start, high at end
Array.Sort(fparray, new FpComparator());
//now for each individual filter result, we count how many we have
//first find out how many different values we have
var numdistinct = 1;
for (var i = 1; i < fparray.Length; i++)
{
if (fparray[i].FilterValue != fparray[i - 1].FilterValue)
{
numdistinct++;
}
}
//now we create an array to hold the filter values and their counts
var results = new double[numdistinct][];
//var results = new double[numdistinct][2];
results[0][0] = fparray[0].FilterValue;
results[0][1] = 1;
var k = 0;
//now we fill up the array
foreach (var t in fparray)
{
if (results[k][0] != t.FilterValue)
{
k++;
results[k][0] = t.FilterValue;
results[k][1] = 1;
}
else
{
results[k][1]++;
}
}
//now normalise the graph
foreach (var t in results)
{
t[1] = t[1] / fparray.Length;
}
//graph produced, return results
return(results);
}
/**
* Gets the laplace graph of an image. The image
* is assumed to be colour, no negative values will
* result.
*
* @param image The image to get the graph of.
* @return The graph of the image.
*/
public static double[][] GetGraph(LockBitmap image)
{
var filter = new Laplace(image, 0, 8);
//filter the image
var fparray =
new FilteredPixel[image.Width*image.Height];
for (var i = 0; i < image.Width; i++)
{
for (var j = 0; j < image.Height; j++)
{
fparray[i*image.Height + j] =
new FilteredPixel(i, j,
Math.Abs(filter.GetValue(i, j)));
}
}
//sort the filter results
//is in ascending order - low at start, high at end
Array.Sort(fparray, new FpComparator());
//now for each individual filter result, we count how many we have
//first find out how many different values we have
var numdistinct = 1;
for (var i = 1; i < fparray.Length; i++)
{
if (fparray[i].FilterValue != fparray[i - 1].FilterValue)
{
numdistinct++;
}
}
//now we create an array to hold the filter values and their counts
//var results = new double[numdistinct][2];
var results = new double[numdistinct][];
for (var i = 0; i < results.Length; i++)
{
results[i] = new double[2];
}
results[0][0] = fparray[0].FilterValue;
results[0][1] = 1;
var k = 0;
//now we fill up the array
foreach (var t in fparray)
{
if (results[k][0] != t.FilterValue)
{
k++;
results[k][0] = t.FilterValue;
results[k][1] = 1;
}
else
{
results[k][1]++;
}
}
//now normalise the graph
foreach (var t in results)
{
t[1] = t[1]/fparray.Length;
}
//graph produced, return results
return results;
}