static void Main()
{
// You probably have to change the path to point to the supplied
// image "river.jpg".
// Because of a bug in Visual studio this project should not be run
// inside the debugger. It will be extremely slow.
Bitmap image = new Bitmap(@"river.jpg");
// Filters can be added and removed here.
Filter[] filters = new Filter[] { Filter.MEDIAN_FILTER, Filter.MEAN_FILTER, Filter.EDGE_DETECTION };
// Do the processing using the pipeline
// Note that the image will be converted to gray scale.
Bitmap result = ImageProcessing.ApplyFiltersPipelined(image, filters, 8);
// ... or use the sequential method
// Bitmap result = ImageProcessing.ApplyFilters(image, filters);
// Show the resulting image
Application.EnableVisualStyles();
Application.SetCompatibleTextRenderingDefault(false);
Application.Run(new ImageForm(new ImagePanel(result)));
// For a text based version uncomment the line below.
// Remember to change the "Output type" of the project
// to "Console application"
// ImageProcessingConsole(image);
}
private static void ImageProcessingConsole(Bitmap image)
{
// we apply a filter 50 times
int numFilters = 50;
// in the pipelined version the image
// is divided into 30 strips
int strips = 30;
Filter[] filters = new Filter[numFilters];
for (int i = 0; i < numFilters; i++)
{
filters[i] = Filter.MEDIAN_FILTER;
}
// Let the JIT compilation kick in
ImageProcessing.ApplyFilters(image, filters);
ImageProcessing.ApplyFiltersPipelined(image, filters, strips);
long start = DateTime.Now.Ticks;
ImageProcessing.ApplyFilters(image, filters);
long end = DateTime.Now.Ticks;
long tSeq = (end - start) / 10000;
Console.WriteLine("Time spent " + tSeq);
start = System.DateTime.Now.Ticks;
ImageProcessing.ApplyFiltersPipelined(image, filters, strips);
end = System.DateTime.Now.Ticks;
long tPar = (end - start) / 10000;
Console.WriteLine("Time spent " + tPar);
Console.WriteLine("Speed up " + (double)tSeq / tPar);
}
