public static void SpeedInt32Test()
{
Console.WriteLine("================ Int Test ============== \n\n");
int count = 10000;
var Fc = new FullConnect <int>(100, 3, 1);
var sig = new Sigmoid <int>();
var net = new Network <int>(OptimiserType.LevenbergMarquardt);
net.Add(Fc);
net.Add(sig);
var rnd = new Random();
var sw = new Stopwatch();
var tInp = new Tensor4 <int>(1, 1, 100, 1, rnd);
Tensor4 <int> outp = new Tensor4 <int>(1, 1, 1, 1);
sw.Start();
for (int j = 0; j < count; j++)
{
outp = net.Forward(tInp);
}
Console.WriteLine("Прямой проход слоя составляет: " + sw.ElapsedMilliseconds / (double)count + " мс");
sw.Stop();
Console.WriteLine("\n\n");
for (int i = 0; i < outp.D; i++)
{
Console.WriteLine(outp[0, 0, i, 0]);
}
}
/// <summary>
/// Reconnectes the Network, remembering the connections
/// between neurons, making FeedNeuralNetwork() faster.
/// </summary>
protected void ConnectNeuralNetwork()
{
if (Network == null)
{
Network = new Dictionary <Neuron, IList <Synapse> >();
}
else
{
Network.Clear();
}
foreach (var gene in Genes)
{
if (!Network.ContainsKey(Neurons[gene.Val.transmitter]))
{
Network.Add(Neurons[gene.Val.transmitter],
new List <Synapse>(1));
}
if (!Network.ContainsKey(Neurons[gene.Val.receiver]))
{
Network.Add(Neurons[gene.Val.receiver],
new List <Synapse>(1));
}
Network[Neurons[gene.Val.transmitter]].Add(gene.Val);
Network[Neurons[gene.Val.receiver]].Add(gene.Val);
}
}
static void Main(string[] args)
{
Network network = new Network();
Layer layer2 = new Layer();
Layer layer3 = new Layer();
Node node1 = new Node(2);
Node node2 = new Node(2);
Node node3 = new Node(2);
layer2.Add(node1);
layer2.Add(node2);
layer3.Add(node3);
network.Add(layer2);
network.Add(layer3);
Dictionary<Signal, Signal> trainingSet = new Dictionary<Signal, Signal>();
Signal signalFF = new Signal();
Signal signalFT = new Signal();
Signal signalTF = new Signal();
Signal signalTT = new Signal();
Signal signalF = new Signal();
Signal signalT = new Signal();
signalFF.Add(0.0); signalFF.Add(0.0);
signalFT.Add(0.0); signalFT.Add(1.0);
signalTF.Add(1.0); signalTF.Add(0.0);
signalTT.Add(1.0); signalTT.Add(1.0);
signalF.Add(0.0);
signalT.Add(1.0);
trainingSet.Add(signalFF, signalF);
trainingSet.Add(signalFT, signalT);
trainingSet.Add(signalTF, signalT);
trainingSet.Add(signalTT, signalF);
network.Train(trainingSet);
Console.WriteLine("OutputFF: " + signalFF * network);
Console.WriteLine("OutputFT: " + signalFT * network);
Console.WriteLine("OutputTF: " + signalTF * network);
Console.WriteLine("OutputTT: " + signalTT * network);
Console.ReadLine();
}
static void Main(string[] args)
{
NetworkElement tunnelNetwork = new Network();
tunnelNetwork.Add(new Tunnel(5));
tunnelNetwork.Add(new Railroad(5));
NetworkElement stationNetwork = new Network();
stationNetwork.Add(new Station());
stationNetwork.Add(new Railroad(10));
NetworkElement bigNetwork = new Network();
bigNetwork.Add(tunnelNetwork);
bigNetwork.Add(stationNetwork);
System.Console.WriteLine(bigNetwork.GetMaintananceCost());
}
public FakeNetworksService()
{
networks = new List <Network>();
Network network = new Network(1, "Network 1", 10);
SubStation substation1 = new SubStation(2, "Substation A");
Device switch1 = new Switch(3, "Switch A");
Line line1 = new Line(4, "Line A", 100);
line1.Terminal1 = substation1;
line1.Terminal2 = switch1;
substation1.Terminal = line1;
switch1.Terminal1 = line1;
SubStation subStation2 = new SubStation(10, "Substation B");
subStation2.Terminal = line1;
switch1.Terminal2 = subStation2;
Load load1 = new Load(5, "Load 1");
line1.Connect(load1);
Load load2 = new Load(7, "Load 2");
line1.Connect(load2);
network.Add(substation1);
network.Add(subStation2);
network.Add(switch1);
network.Add(line1);
networks.Add(network);
}
private void Add()
{
using (LayerDialog dialog = new LayerDialog()
{
Layer = (Layer)TemplateLayer.Clone()
})
{
if (dialog.ShowDialog(this) == DialogResult.OK)
{
Network.Add(dialog.Layer);
UpdateItems();
}
}
}
/// <summary>
/// This is the method that actually does the work.
/// </summary>
/// <param name="DA">The DA object can be used to retrieve data from input parameters and
/// to store data in output parameters.</param>
protected override void SolveInstance(IGH_DataAccess DA)
{
string path = "";
int size = 0;
bool debug = false;
int maxNgon = 3;
double tol = -1;
if (!DA.GetData(0, ref path))
{
return;
}
if (!DA.GetData(1, ref size))
{
return;
}
DA.GetData(2, ref maxNgon);
DA.GetData(3, ref tol);
IplImage img = new IplImage(path, LoadMode.Color);
IplImage imgHSV = new IplImage(img.Size, BitDepth.U8, 3);
Cv.CvtColor(img, imgHSV, ColorConversion.RgbToHsv);
var channels = imgHSV.Split();
IplImage hue = channels[0];
IplImage Render = new IplImage(img.Size, BitDepth.U8, 3);
//色抽出マスク用画像宣言
IplImage imgB1 = new IplImage(img.Size, BitDepth.U8, 1);
IplImage imgB2 = new IplImage(img.Size, BitDepth.U8, 1);
IplImage imgR = new IplImage(img.Size, BitDepth.U8, 1);
IplImage imgG = new IplImage(img.Size, BitDepth.U8, 1);
IplImage imgB = new IplImage(img.Size, BitDepth.U8, 1);
int RG = 30;
int GB = 90;
int BR = 150;
int off = 1;
int smin = 30;
int bmin = 30;
//色抽出用閾値作成
CvScalar Bmin1 = new CvScalar(0, smin, bmin);
CvScalar Bmax1 = new CvScalar(RG - off, 255, 255);
CvScalar Bmin2 = new CvScalar(BR + off, smin, bmin);
CvScalar Bmax2 = new CvScalar(180, 255, 255);
CvScalar Gmin = new CvScalar(RG + off, smin, bmin);
CvScalar Gmax = new CvScalar(GB - off, 255, 255);
CvScalar Rmin = new CvScalar(GB + off, smin, bmin);
CvScalar Rmax = new CvScalar(BR - off, 255, 255);
//閾値を用いて色抽出
Cv.InRangeS(imgHSV, Bmin1, Bmax1, imgB1);
Cv.InRangeS(imgHSV, Bmin2, Bmax2, imgB2);
Cv.Add(imgB1, imgB2, imgB);
Cv.InRangeS(imgHSV, Gmin, Gmax, imgG);
Cv.InRangeS(imgHSV, Rmin, Rmax, imgR);
//Blobs化
CvBlobs Rs = new CvBlobs(imgR);
CvBlobs Gs = new CvBlobs(imgG);
CvBlobs Bs = new CvBlobs(imgB);
int minArea = img.Width * img.Height / 20000;
int maxArea = img.Width * img.Height;
Bs.FilterByArea(minArea, maxArea);
Rs.FilterByArea(minArea, maxArea);
Gs.FilterByArea(minArea, maxArea);
//blobの配列化
CvBlob[] Rblobs = new CvBlob[Rs.Count];
CvBlob[] Bblobs = new CvBlob[Bs.Count];
CvBlob[] Gblobs = new CvBlob[Gs.Count];
Rs.Values.CopyTo(Rblobs, 0);
Bs.Values.CopyTo(Bblobs, 0);
Gs.Values.CopyTo(Gblobs, 0);
if (!debug)
{
string deb = "";
foreach (var bbbb in Rblobs)
{
deb += bbbb.Area + "\r\n";
}
//BlobからLine化
List <Line> Rlines = ExtractLinesFromBlobs(Rblobs);
List <Line> Blines = ExtractLinesFromBlobs(Bblobs);
List <Line> Glines = ExtractLinesFromBlobs(Gblobs);
//scale
double MinSize = Math.Min(img.Width, img.Height);
double ScaleFactor = (double)size / MinSize;
var scale = Transform.Scale(new Point3d(0, 0, 0), ScaleFactor);
Network network = new Network();
//ネットワークにLineを色ごとにラベル付きで入れる
foreach (var l in Rlines)
{
l.Transform(scale);
network.Add(l, 0);
}
foreach (var l in Blines)
{
l.Transform(scale);
network.Add(l, 1);
}
foreach (var l in Glines)
{
l.Transform(scale);
network.Add(l, 2);
}
double t = network.SearchWeldToleranceBinary(0, (double)size / 10, 0, 10);
if (tol != -1)
{
network.weld(tol * size);
}
else
{
network.weld(t);
}
deb += "tolerance: " + t + "\r\n\r\n";
//ウェルド後のエッジ抽出
Rlines = network.ExtractLines(0);
Blines = network.ExtractLines(1);
Glines = network.ExtractLines(2);
List <List <int> > faces = network.detectCycles(maxNgon);
deb += "B: " + Bs.Count.ToString() + "\r\n";
deb += "R: " + Rs.Count.ToString() + "\r\n";
deb += "G: " + Gs.Count.ToString() + "\r\n";
Mesh mesh = GenerateMesh(network.verts, faces);
mesh.Normals.ComputeNormals();
DA.SetDataList(0, network.verts);
DA.SetDataList(1, Rlines);
DA.SetDataList(2, Blines);
DA.SetDataList(3, Glines);
DA.SetData(4, mesh);
}
else
{
//赤レンダリング
Rs.RenderBlobs(img, Render, RenderBlobsMode.Angle);
Rs.RenderBlobs(img, Render, RenderBlobsMode.BoundingBox);
Rs.RenderBlobs(img, Render, RenderBlobsMode.Centroid);
//青レンダリング
Bs.RenderBlobs(img, Render, RenderBlobsMode.Angle);
Bs.RenderBlobs(img, Render, RenderBlobsMode.BoundingBox);
Bs.RenderBlobs(img, Render, RenderBlobsMode.Centroid);
//黒レンダリング
Gs.RenderBlobs(img, Render, RenderBlobsMode.Angle);
Gs.RenderBlobs(img, Render, RenderBlobsMode.BoundingBox);
Gs.RenderBlobs(img, Render, RenderBlobsMode.Centroid);
Cv.NamedWindow("test");
IplImage Render2 = new IplImage(img.Size.Width / 4, img.Size.Height / 4, BitDepth.U8, 3);
string deb = "";
deb += "B: " + Bs.Count.ToString() + "\r\n";
deb += "R: " + Rs.Count.ToString() + "\r\n";
deb += "G: " + Gs.Count.ToString() + "\r\n";
Cv.Resize(Render, Render2);
Cv.ShowImage("test", Render2);
Cv.WaitKey();
Cv.DestroyWindow("test");
}
Cv.ReleaseImage(img);
Cv.ReleaseImage(imgHSV);
}
