private static int DetectPeleeNet(NcnnDotNet.OpenCV.Mat bgr, List <Object> objects, NcnnDotNet.Mat resized)
{
using (var peleenet = new Net())
{
if (Ncnn.IsSupportVulkan)
{
peleenet.Opt.UseVulkanCompute = true;
}
// model is converted from https://github.com/eric612/MobileNet-YOLO
// and can be downloaded from https://drive.google.com/open?id=1Wt6jKv13sBRMHgrGAJYlOlRF-o80pC0g
// the ncnn model https://github.com/nihui/ncnn-assets/tree/master/models
peleenet.LoadParam("pelee.param");
peleenet.LoadModel("pelee.bin");
const int targetSize = 300;
var imgW = bgr.Cols;
var imgH = bgr.Rows;
using var @in = Mat.FromPixelsResize(bgr.Data, PixelType.Bgr, bgr.Cols, bgr.Rows, targetSize, targetSize);
var meanVals = new[] { 103.9f, 116.7f, 123.6f };
var normVals = new[] { 0.017f, 0.017f, 0.017f };
@in.SubstractMeanNormalize(meanVals, normVals);
using var ex = peleenet.CreateExtractor();
//ex.SetNumThreads(4);
ex.Input("data", @in);
using var @out = new Mat();
ex.Extract("detection_out", @out);
// printf("%d %d %d\n", out.w, out.h, out.c);
objects.Clear();
for (var i = 0; i < @out.H; i++)
{
var values = @out.Row(i);
var @object = new Object();
@object.Label = (int)values[0];
@object.Prob = values[1];
@object.Rect.X = values[2] * imgW;
@object.Rect.Y = values[3] * imgH;
@object.Rect.Width = values[4] * imgW - @object.Rect.X;
@object.Rect.Height = values[5] * imgH - @object.Rect.Y;
objects.Add(@object);
}
using var segOut = new Mat();
ex.Extract("sigmoid", segOut);
Ncnn.ResizeBilinear(segOut, resized, imgW, imgH);
}
return(0);
}
private int Detect(Mat image,
int resizedWidth,
int resizedHeight,
float scoreThreshold,
float nmsThreshold,
List <FaceInfo> faces)
{
if (image.IsEmpty)
{
Console.WriteLine("image is empty ,please check!");
return(-1);
}
this._ImageH = image.H;
this._ImageW = image.W;
this._ScaleW = (float)this._ImageW / resizedWidth;
this._ScaleH = (float)this._ImageH / resizedHeight;
using (Mat @in = new Mat())
{
//scale
this.DynamicScale(resizedWidth, resizedHeight);
Ncnn.ResizeBilinear(image, @in, this._DW, this._DH);
using (Extractor ex = this._Net.CreateExtractor())
{
ex.Input("input.1", @in);
using (Mat heatMap = new Mat())
{
using (Mat scale = new Mat())
{
using (Mat offset = new Mat())
{
using (Mat landmarks = new Mat())
{
ex.Extract("537", heatMap);
ex.Extract("538", scale);
ex.Extract("539", offset);
ex.Extract("540", landmarks);
this.Decode(heatMap, scale, offset, landmarks, faces, scoreThreshold, nmsThreshold);
this.SquareBox(faces);
}
}
}
}
}
}
return(0);
}
private int Detect(Mat img, ICollection <FaceInfo> faceList)
{
if (img.IsEmpty)
{
Console.WriteLine("image is empty ,please check!");
return(-1);
}
this._ImageH = img.H;
this._ImageW = img.W;
using (var @in = new Mat())
{
Ncnn.ResizeBilinear(img, @in, this._InW, this._InH);
var ncnnImg = @in;
ncnnImg.SubstractMeanNormalize(this._MeanVals, this._NormVals);
var boundingBoxCollection = new List <FaceInfo>();
//var validInput = new List<FaceInfo>();
using (var ex = this._UltraFace.CreateExtractor())
{
ex.SetNumThreads(this._NumThread);
ex.Input("input", ncnnImg);
using (var scores = new Mat())
using (var boxes = new Mat())
{
ex.Extract("scores", scores);
ex.Extract("boxes", boxes);
GenerateBBox(boundingBoxCollection, scores, boxes, this._ScoreThreshold, this._NumAnchors);
NonMaximumSuppression(boundingBoxCollection, faceList);
}
}
}
return(0);
}