ImageResizer.Plugins.Faces.FaceDetection Class Reference

Provides a simple API for face detection More...

Inheritance diagram for ImageResizer.Plugins.Faces.FaceDetection:

Collaboration diagram for ImageResizer.Plugins.Faces.FaceDetection:

Protected Member Functions
override List< Face >	DetectFeatures (IplImage img, CvMemStorage storage)
	Detects features on a grayscale image. More...
Protected Member Functions inherited from ImageResizer.Plugins.Faces.FeatureDetectionBase< T >
int	CompareByNeighbors (CvAvgComp a, CvAvgComp b)

Properties
int	MinFaces haarcascade_frontalface_default.xml" [get, set]
	The minimum number of faces expected More...
int	MaxFaces [get, set]
	The maximum number of faces wanted More...
float	MinSizePercent [get, set]
	The smallest face that will be detected, represented in a 1..100 percentage of the (smaller of width and height). Defaults to 3 percent (3.0f) More...
int	ConfidenceLevelThreshold [get, set]
	The minimum number of agreeing matches required for a face rectangle to be returned. This rule isn't applied if we don't have [MinFaces] number of faces. More...
int	MinConfidenceLevel [get, set]
	The minimum number of agreeing matches required for a face rectangle to be evaluated More...
double	ExpandX [get, set]
	The percentage by which to expand each face rectangle horizontally after detection. To expand 5% each side, set to 0.1 More...
double	ExpandY [get, set]
	The percentage by which to expand each face rectangle vertically after detection. To expand 20% on the top and bottom, set to 0.4 More...

Additional Inherited Members
Public Member Functions inherited from ImageResizer.Plugins.Faces.FeatureDetectionBase< T >
	FeatureDetectionBase (string xmlFolder)
List< T >	DetectFeatures (Bitmap b)
void	Dispose ()
	Disposes all loaded cascades More...
Protected Attributes inherited from ImageResizer.Plugins.Faces.FeatureDetectionBase< T >
List< string >	searchFolders = new List<string>() { }
Dictionary< string, string >	fileNames = new Dictionary<string,string>(){ }
Dictionary< string, string >	Files = null
Dictionary< string, CvHaarClassifierCascade >	Cascades = null
int	scaledBounds = 1000
	Large images will be scaled down to less than scaledBounds X scaledBounds for feature detection. Defaults to 1000 More...

Detailed Description

Provides a simple API for face detection

Definition at line 30 of file FaceDetection.cs.

Member Function Documentation

override List<Face> ImageResizer.Plugins.Faces.FaceDetection.DetectFeatures ( IplImage  img, CvMemStorage  storage  )

inlineprotectedvirtual

Detects features on a grayscale image.

Parameters

img
storage

Returns

Implements ImageResizer.Plugins.Faces.FeatureDetectionBase< T >.

Definition at line 87 of file FaceDetection.cs.

                                                                                         {
            //Determine minimum face size
            var minSize = (int)Math.Round((double)MinSizePercent / 100.0 * Math.Min(img.Width, img.Height));

            
            //Detect faces (frontal). TODO: side
            Stopwatch watch = Stopwatch.StartNew();
            CvAvgComp[] faces = Cv.HaarDetectObjects(img, Cascades["FaceCascade"], storage, 1.0850, MinConfidenceLevel, 0, new CvSize(minSize, minSize)).ToArrayAndDispose();
            watch.Stop();
            Debug.WriteLine("Face detection time = " + watch.ElapsedMilliseconds);

            //Sort by accuracy
            Array.Sort<CvAvgComp>(faces, CompareByNeighbors);

            //Convert into feature objects list
            List<Face> features = new List<Face>(faces.Length);
            foreach (CvAvgComp face in faces) features.Add(new Face(PolygonMath.ScaleRect(face.Rect.ToRectangleF(),ExpandX,ExpandY), face.Neighbors));
            
            //Unless we're below MinFaces, filter out the low confidence matches.
            while (features.Count > MinFaces && features[features.Count - 1].Accuracy < ConfidenceLevelThreshold) features.RemoveAt(features.Count - 1);


            //Never return more than [MaxFaces]
            return (features.Count > MaxFaces) ? features.GetRange(0, MaxFaces) : features;
        }

Property Documentation

int ImageResizer.Plugins.Faces.FaceDetection.ConfidenceLevelThreshold

getset

The minimum number of agreeing matches required for a face rectangle to be returned. This rule isn't applied if we don't have [MinFaces] number of faces.

Definition at line 63 of file FaceDetection.cs.

double ImageResizer.Plugins.Faces.FaceDetection.ExpandX

getset

The percentage by which to expand each face rectangle horizontally after detection. To expand 5% each side, set to 0.1

Definition at line 73 of file FaceDetection.cs.

double ImageResizer.Plugins.Faces.FaceDetection.ExpandY

getset

The percentage by which to expand each face rectangle vertically after detection. To expand 20% on the top and bottom, set to 0.4

Definition at line 78 of file FaceDetection.cs.

int ImageResizer.Plugins.Faces.FaceDetection.MaxFaces

getset

The maximum number of faces wanted

Definition at line 51 of file FaceDetection.cs.

int ImageResizer.Plugins.Faces.FaceDetection.MinConfidenceLevel

getset

The minimum number of agreeing matches required for a face rectangle to be evaluated

Definition at line 68 of file FaceDetection.cs.

int ImageResizer.Plugins.Faces.FaceDetection.MinFaces haarcascade_frontalface_default.xml"

getset

The minimum number of faces expected

Definition at line 47 of file FaceDetection.cs.

float ImageResizer.Plugins.Faces.FaceDetection.MinSizePercent

getset

The smallest face that will be detected, represented in a 1..100 percentage of the (smaller of width and height). Defaults to 3 percent (3.0f)

Definition at line 57 of file FaceDetection.cs.

---

The documentation for this class was generated from the following file:

• C:/Users/nathanael/Documents/resizer/Plugins/Faces/FaceDetection.cs