/// <summary>
/// 活体检测器。
/// <para>
/// 视频帧图片，由 <paramref name="global"/> 指定是否启用全局检测能力 <br />
/// </para>
/// <para>如果返回结果为 <see cref="AntiSpoofingStatus.Detecting"/>，则说明需要继续调用此方法，传入更多的图片</para>
/// </summary>
/// <param name="viewFace"></param>
/// <param name="bitmaps">一组图片，即视频帧的 <see cref="Bitmap"/> 数组</param>
/// <param name="faceIndex">指定要识别的人脸索引</param>
/// <param name="global">是否启用全局检测能力</param>
/// <returns></returns>
public static AntiSpoofingStatus AntiSpoofingVideo(this ViewFace viewFace, Bitmap[] bitmaps, int faceIndex, bool global)
{
var result = AntiSpoofingStatus.Detecting;
bool haveFace = false;
foreach (var bitmap in bitmaps)
{
var infos = viewFace.FaceDetector(bitmap);
if (faceIndex >= 0 && faceIndex < infos.Length)
{
haveFace = true;
var points = viewFace.FaceMark(bitmap, infos[faceIndex]);
var status = viewFace.AntiSpoofingVideo(bitmap, infos[faceIndex], points, global);
if (status == AntiSpoofingStatus.Detecting)
{
continue;
}
else
{
result = status;
}
}
}
if (haveFace)
{
return(result);
}
else
{
return(AntiSpoofingStatus.Error);
}
}
public void TestAntiSpoofingVideo(bool global)
{
if (!(Image.FromFile(images[0]) is Bitmap bitmap))
{
throw new ArgumentNullException(nameof(bitmap));
}
ViewFace viewFace = new ViewFace((str) => { System.Console.WriteLine(str); });
var infos = viewFace.FaceDetector(bitmap);
if (infos.Length > 0)
{
var points = viewFace.FaceMark(bitmap, infos[0]);
AntiSpoofingStatus status = AntiSpoofingStatus.Error;
while (true)
{
status = viewFace.AntiSpoofingVideo(bitmap, infos[0], points, global);
if (status != AntiSpoofingStatus.Detecting)
{
break;
}
}
Console.WriteLine(status);
}
else
{
Console.WriteLine("No Face!");
}
bitmap.Dispose();
}
/// <summary>
/// 计算人脸特征值相似度。
/// </summary>
/// <param name="viewFace"></param>
/// <param name="lhs"></param>
/// <param name="rhs"></param>
/// <returns></returns>
public static float Compare(this ViewFace viewFace, float[] lhs, float[] rhs)
{
float sum = 0;
for (int i = 0; i < lhs.Length; i++)
{
sum += lhs[i] * rhs[i];
}
return(sum);
}
static void Main()
{
ViewFace viewFace = new ViewFace((str) => { Debug.WriteLine(str); }); // 初始化人脸识别类，并设置 日志回调函数
viewFace.DetectorSetting = new DetectorSetting()
{
FaceSize = 20, MaxWidth = 2000, MaxHeight = 2000, Threshold = 0.5
};
// 系统默认使用的轻量级识别模型。如果对精度有要求，请切换到 Normal 模式；并下载需要模型文件 放入生成目录的 model 文件夹中
viewFace.FaceType = FaceType.Normal;
// 系统默认使用5个人脸关键点。//不建议改动，除非是使用口罩模型。
viewFace.MarkType = MarkType.Light;
#region 识别老照片
float[] oldEigenValues;
Bitmap oldImg = (Bitmap)Image.FromFile(@"C:\Users\yangw\OneDrive\图片\Camera Roll\IMG_20181103_142707.jpg" /*老图片路径*/); // 从文件中加载照片 // 或者视频帧等
var oldFaces = viewFace.FaceDetector(oldImg); // 检测图片中包含的人脸信息。(置信度、位置、大小)
if (oldFaces.Length > 0) //识别到人脸
{
{ // 打印人脸信息
Console.WriteLine($"识别到的人脸数量：{oldFaces.Length} 。人脸信息：\n");
Console.WriteLine($"序号\t人脸置信度\t位置X\t位置Y\t宽度\t高度");
for (int i = 0; i < oldFaces.Length; i++)
{
Console.WriteLine($"{i + 1}\t{oldFaces[i].Score}\t{oldFaces[i].Location.X}\t{oldFaces[i].Location.Y}\t{oldFaces[i].Location.Width}\t{oldFaces[i].Location.Height}");
}
Console.WriteLine();
}
var oldPoints = viewFace.FaceMark(oldImg, oldFaces[0]); // 获取 第一个人脸 的识别关键点。(人脸识别的关键点数据)
oldEigenValues = viewFace.Extract(oldImg, oldPoints); // 获取 指定的关键点 的特征值。
}
else
{
oldEigenValues = new float[0]; /*未识别到人脸*/
}
#endregion
#region 识别新照片
float[] newEigenValues;
Bitmap newImg = (Bitmap)Image.FromFile(@"C:\Users\yangw\OneDrive\图片\Camera Roll\IMG_20181129_224339.jpg" /*新图片路径*/); // 从文件中加载照片 // 或者视频帧等
var newFaces = viewFace.FaceDetector(newImg); // 检测图片中包含的人脸信息。(置信度、位置、大小)
if (newFaces.Length > 0) //识别到人脸
{
{ // 打印人脸信息
Console.WriteLine($"识别到的人脸数量：{newFaces.Length} 。人脸信息：\n");
Console.WriteLine($"序号\t人脸置信度\t位置X\t位置Y\t宽度\t高度");
for (int i = 0; i < newFaces.Length; i++)
{
Console.WriteLine($"{i + 1}\t{newFaces[i].Score}\t{newFaces[i].Location.X}\t{newFaces[i].Location.Y}\t{newFaces[i].Location.Width}\t{newFaces[i].Location.Height}");
}
Console.WriteLine();
}
var newPoints = viewFace.FaceMark(newImg, newFaces[0]); // 获取 第一个人脸 的识别关键点。(人脸识别的关键点数据)
newEigenValues = viewFace.Extract(newImg, newPoints); // 获取 指定的关键点 的特征值。
}
else
{
newEigenValues = new float[0]; /*未识别到人脸*/
}
#endregion
try
{
float similarity = viewFace.Similarity(oldEigenValues, newEigenValues); // 对比两张照片上的数据，确认是否是同一个人。
Console.WriteLine($"阈值 = {Face.Threshold[viewFace.FaceType]}\t相似度 = {similarity}");
Console.WriteLine($"是否是同一个人：{viewFace.IsSelf(similarity)}");
}
catch (Exception e)
{ Console.WriteLine(e); }
Console.ReadKey();
}
static void Main()
{
// 老图片路径
string oldImgPath = @"C:/Users/yangw/OneDrive/桌面/image/0.jpg";
string newImgPath = @"C:/Users/yangw/OneDrive/桌面/image/1.jpg";
ViewFace viewFace = new ViewFace((str) => { Debug.WriteLine(str); }); // 初始化人脸识别类，并设置 日志回调函数
// 系统默认使用的轻量级识别模型。如果对精度有要求，请切换到 Normal 模式；并下载需要模型文件 放入生成目录的 model 文件夹中
viewFace.FaceType = FaceType.Normal;
// 系统默认使用5个人脸关键点。//不建议改动，除非是使用口罩模型。
viewFace.MarkType = MarkType.Light;
#region 识别老照片
float[] oldEigenValues;
Bitmap oldImg = (Bitmap)Image.FromFile(oldImgPath); // 从文件中加载照片 // 或者视频帧等
var oldFaceInfos = viewFace.FaceDetector(oldImg); // 检测图片中包含的人脸信息。(置信度、位置、大小)
oldFaceInfos = viewFace.FaceDetectorAsync(oldImg).GetAwaiter().GetResult(); // 异步检测图片中的人脸信息
if (oldFaceInfos.Length > 0) //识别到人脸
{
{ // 打印人脸信息
Console.WriteLine($"识别到的人脸数量：{oldFaceInfos.Length} 。人脸信息：\n");
Console.WriteLine($"序号\t人脸置信度\t位置X\t位置Y\t宽度\t高度");
for (int i = 0; i < oldFaceInfos.Length; i++)
{
Console.WriteLine($"{i + 1}\t{oldFaceInfos[i].Score:f8}\t{oldFaceInfos[i].Location.X}\t{oldFaceInfos[i].Location.Y}\t{oldFaceInfos[i].Location.Width}\t{oldFaceInfos[i].Location.Height}");
}
Console.WriteLine();
}
var oldPoints = viewFace.FaceMark(oldImg, oldFaceInfos[0]); // 获取 第一个人脸 的识别关键点。(人脸识别的关键点数据)
oldEigenValues = viewFace.Extract(oldImg, oldPoints); // 获取 指定的关键点 的特征值。
var state = viewFace.AntiSpoofing(oldImg, oldFaceInfos[0], oldPoints);
Console.WriteLine($"单帧 活体检测结果：{state}");
state = viewFace.AntiSpoofingVideo(oldImg, oldFaceInfos[0], oldPoints);
Console.WriteLine($"视频 活体检测结果：{state}");
Console.WriteLine();
#region 质量评估
var qualityBrightness = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.Brightness);
Console.WriteLine($"亮度:Level [{qualityBrightness.Level}] - Score [{qualityBrightness.Score}]");
var qualityClarity = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.Clarity);
Console.WriteLine($"清晰度:Level [{qualityClarity.Level}] - Score [{qualityClarity.Score}]");
var qualityIntegrity = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.Integrity);
Console.WriteLine($"完整度:Level [{qualityIntegrity.Level}] - Score [{qualityIntegrity.Score}]");
var qualityPose = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.Pose);
Console.WriteLine($"姿态:Level [{qualityPose.Level}] - Score [{qualityPose.Score}]");
var qualityPoseEx = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.PoseEx);
Console.WriteLine($"姿态 (深度):Level [{qualityPoseEx.Level}] - Score [{qualityPoseEx.Score}]");
var qualityResolution = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.Resolution);
Console.WriteLine($"分辨率:Level [{qualityResolution.Level}] - Score [{qualityResolution.Score}]");
var qualityStructure = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.Structure);
Console.WriteLine($"遮挡:Level [{qualityStructure.Level}] - Score [{qualityStructure.Score}]");
var qualityClarityEx = viewFace.FaceQuality(oldImg, oldFaceInfos[0], oldPoints, QualityType.ClarityEx);
Console.WriteLine($"清晰度 (深度):Level [{qualityClarityEx.Level}] - Score [{qualityClarityEx.Score}]");
#endregion
}
else
{
oldEigenValues = new float[0]; /*未识别到人脸*/
}
#endregion
Console.WriteLine();
#region 人脸追踪
var trackFaceInfos = viewFace.FaceTrack(oldImg); // 人脸追踪
if (trackFaceInfos.Length > 0)
{
{ // 打印人脸信息
Console.WriteLine($"\r\n跟踪到的人脸数量：{trackFaceInfos.Length} 。人脸信息：\n");
Console.WriteLine($"PID\t人脸置信度\t位置X\t位置Y\t宽度\t高度");
for (int i = 0; i < trackFaceInfos.Length; i++)
{
Console.WriteLine($"{trackFaceInfos[i].Pid}\t{trackFaceInfos[i].Score:f8}\t{trackFaceInfos[i].Location.X}\t{trackFaceInfos[i].Location.Y}\t{trackFaceInfos[i].Location.Width}\t{trackFaceInfos[i].Location.Height}");
}
}
}
#endregion
Console.WriteLine();
#region 识别新照片
float[] newEigenValues;
Bitmap newImg = (Bitmap)Image.FromFile(newImgPath /*新图片路径*/); // 从文件中加载照片 // 或者视频帧等
newImg = (Bitmap)newImg.ChangeSize(new Size(1024, 768));
var newFaces = viewFace.FaceDetector(newImg); // 检测图片中包含的人脸信息。(置信度、位置、大小)
if (newFaces.Length > 0) //识别到人脸
{
{ // 打印人脸信息
Console.WriteLine($"识别到的人脸数量：{newFaces.Length} 。人脸信息：\n");
Console.WriteLine($"序号\t人脸置信度\t位置X\t位置Y\t宽度\t高度");
for (int i = 0; i < newFaces.Length; i++)
{
Console.WriteLine($"{i + 1}\t{newFaces[i].Score:f8}\t{newFaces[i].Location.X}\t{newFaces[i].Location.Y}\t{newFaces[i].Location.Width}\t{newFaces[i].Location.Height}");
}
}
var newPoints = viewFace.FaceMark(newImg, newFaces[0]); // 获取 第一个人脸 的识别关键点。(人脸识别的关键点数据)
newEigenValues = viewFace.Extract(newImg, newPoints); // 获取 指定的关键点 的特征值。
}
else
{
newEigenValues = new float[0]; /*未识别到人脸*/
}
#endregion
Console.WriteLine();
try
{
float similarity = viewFace.Similarity(oldEigenValues, newEigenValues); // 对比两张照片上的数据，确认是否是同一个人。
Console.WriteLine($"阈值 = {FaceCompareConfig.GetThreshold(viewFace.FaceType)}\t相似度 = {similarity}");
Console.WriteLine($"是否是同一个人：{viewFace.IsSelf(similarity)}");
}
catch (Exception e)
{ Console.WriteLine(e.Message); }
Console.ReadKey();
}
