public async Task <string> FaceDetect(string filePath)
{
return(await Task.Run(() =>
{
try
{
var image = Convert.ToBase64String(File.ReadAllBytes(filePath));
var imageType = "BASE64";
// 调用人脸检测，可能会抛出网络等异常，请使用try/catch捕获
var result = _client.Detect(image, imageType);
Console.WriteLine(result);
// 如果有可选参数
//var options = new Dictionary<string, object>
//{
// {"face_field", "age"},
// {"max_face_num", 2},
// {"face_type", "LIVE"},
// {"liveness_control", "LOW"}
//};
//// 带参数调用人脸检测
//result = _client.Detect(image, imageType, options);
return result.ToString();
}
catch (Exception e)
{
_logger.LogError(null, e);
throw;
}
}));
}
/// <summary>
/// 人脸检测
/// 检测请求图片中的人脸，返回人脸位置、72个关键点坐标、及人脸相关属性信息。检测响应速度，与图片中人脸数量相关，人脸数量较多时响应时间会有些许延长。
/// </summary>
/// <param name="ImageFilePath">图片文件路径</param>
/// <param name="FaceDetectOptions">默认只返回人脸框、概率和旋转角度。</param>
/// <param name="max_face_num">检测人脸数量，默认为1</param>
/// <returns>
/// 字段 是否必选 类型 说明
/// log_id number 是 日志id
/// result_num number 是 人脸数目
///result array 是 人脸属性对象的集合
///+age number 否 年龄。face_fields包含age时返回
///+beauty number 否 美丑打分，范围0-100，越大表示越美。face_fields包含beauty时返回
///+location object 是 人脸在图片中的位置
///++left number 是 人脸区域离左边界的距离
///++top number 是 人脸区域离上边界的距离
///++width number 是 人脸区域的宽度
///++height number 是 人脸区域的高度
///+face_probability number 是 人脸置信度，范围0-1
///+rotation_angle number 是 人脸框相对于竖直方向的顺时针旋转角，[-180,180]
///+yaw number 是 三维旋转之左右旋转角[-90(左), 90(右)]
///+pitch number 是 三维旋转之俯仰角度[-90(上), 90(下)]
///+roll number 是 平面内旋转角[-180(逆时针), 180(顺时针)]
///+expression number 否 表情，0，不笑；1，微笑；2，大笑。face_fields包含expression时返回
///+expression_probability number 否 表情置信度，范围0 ~1。face_fields包含expression时返回
///+faceshape array 否 脸型置信度。face_fields包含faceshape时返回
///++type string 是 脸型：square/triangle/oval/heart/round
///++probability number 是 置信度：0~1
///+gender string 否 male、female。face_fields包含gender时返回
///+gender_probability number 否 性别置信度，范围0 ~1。face_fields包含gender时返回
///+glasses number 否 是否带眼镜，0-无眼镜，1-普通眼镜，2-墨镜。face_fields包含glasses时返回
///+glasses_probability number 否 眼镜置信度，范围0 ~1。face_fields包含glasses时返回
///+landmark array 否 4个关键点位置，左眼中心、右眼中心、鼻尖、嘴中心。face_fields包含landmark时返回
///++x number 否 x坐标
///++y number 否 y坐标
///+landmark72 array 否 72个特征点位置，示例图 。face_fields包含landmark时返回
///++x number 否 x坐标
///++y number 否 y坐标
///+race string 否 yellow、white、black、arabs。face_fields包含race时返回
///+race_probability number 否 人种置信度，范围0 ~1。face_fields包含race时返回
///+qualities object 否 人脸质量信息。face_fields包含qualities时返回
///++occlusion object 是 人脸各部分遮挡的概率,[0, 1],0表示完整，1表示不完整
///+++left_eye number 是 左眼
///+++right_eye number 是 右眼
///+++nose number 是 鼻子
///+++mouth number 是 嘴
///+++left_cheek number 是 左脸颊
///+++right_cheek number 是 右脸颊
///+++chin number 是 下巴
///++blur number 是 人脸模糊程度，[0, 1]。0表示清晰，1表示模糊
///++illumination number 是 取值范围在[0, 255], 表示脸部区域的光照程度
///++completeness number 是 人脸完整度，0或1, 0为人脸溢出图像边界，1为人脸都在图像边界内
///++type object 是 真实人脸/卡通人脸置信度
///+++human number 是 真实人脸置信度，[0, 1]
///+++cartoon number 是 卡通人脸置信度，[0, 1]
///
///
///可通过人脸检测接口，基于以下字段和对应阈值，进行质量检测的判断，以保证人脸质量符合后续业务操作要求。
///
///遮挡范围
///occlusion（0~1），0为无遮挡，1是完全遮挡含有多个具体子字段，表示脸部多个部位通常用作判断头发、墨镜、口罩等遮挡
///left_eye : 0.6, #左眼被遮挡的阈值
///right_eye : 0.6, #右眼被遮挡的阈值
///nose : 0.7, #鼻子被遮挡的阈值
///mouth : 0.7, #嘴巴被遮挡的阈值
///left_check : 0.8, #左脸颊被遮挡的阈值
///right_check : 0.8, #右脸颊被遮挡的阈值
///chin_contour : 0.6, #下巴被遮挡阈值
///
///模糊度范围
///Blur（0~1），0是最清晰，1是最模糊
///小于0.7
///
///光照范围
///illumination（0~255）脸部光照的灰度值，0表示光照不好以及对应客户端SDK中，YUV的Y分量
///大于40
///
///姿态角度
///Pitch：三维旋转之俯仰角度[-90(上), 90(下)]
///Roll：平面内旋转角[-180(逆时针), 180(顺时针)]
///Yaw：三维旋转之左右旋转角[-90(左), 90(右)]
///分别小于20度
///
///人脸完整度
///completeness（0~1），0代表完整，1代表不完整
///小于0.4
///
///人脸大小
///人脸部分的大小 建议长宽像素值范围：80*80~200*200
///人脸部分不小于100*100像素
/// </returns>
public static BaseResponse FaceDetect(byte[] imageByte, List <FaceDetectOptions> FaceDetectOptions, int max_face_num = 1)
{
if (imageByte.Length > 1024 * 1024 * 10)
{
throw new Exception("图片大小必须小于10Mb");
}
var imageBase64 = Convert.ToBase64String(imageByte);
string face_fields_options = "";
foreach (var item in FaceDetectOptions)
{
face_fields_options += "," + item.ToString();
}
if (face_fields_options.Length != 0)
{
face_fields_options = face_fields_options.Remove(0, 1);
}
var options = new Dictionary <string, object>()
{
{ "face_fields", face_fields_options },
{ "max_face_num", max_face_num }
};
return(Execute(client.Detect(imageBase64, ImageType, options)));
}
/// <summary>
/// 人脸识别
/// </summary>
/// <param name="image">人脸图片</param>
/// <returns></returns>
public JObject NetRecognition(Image image)
{
try
{
var client = new Baidu.Aip.Face.Face(API_KEY, SECRET_KEY)
{
Timeout = 60000 // 修改超时时间
};
//图片转为Base64
Bitmap bmp = new Bitmap(image);
MemoryStream ms = new MemoryStream();
bmp.Save(ms, System.Drawing.Imaging.ImageFormat.Jpeg);
byte[] arr = new byte[ms.Length]; ms.Position = 0;
ms.Read(arr, 0, (int)ms.Length); ms.Close();
var base64 = Convert.ToBase64String(arr);
string imageType = "BASE64";
//可选参数
var options = new Dictionary <string, object> {
{ "face_field", "age,beauty,gender" },
{ "max_face_num", 1 },
{ "face_type", "LIVE" }
};
JObject result = client.Detect(base64, imageType, options);
return(result);
}
catch (Exception ex)
{
MessageBox.Show("连接人脸识别API出错：" + ex);
return(new JObject());
}
}
protected static void Detect()
{
// 设置APPID/AK/SK
var client = new Baidu.Aip.Face.Face(API_KEY, SECRET_KEY);
client.Timeout = 60000; // 修改超时时间
var image = ComFunc.StreamToBase64String(new FileStream("C:/Users/admin/Desktop/微信图片_20190104172518.jpg", FileMode.Open));
var imageType = "BASE64";
// 调用人脸检测，可能会抛出网络等异常，请使用try/catch捕获
//var result = client.Detect(image, imageType);
//Console.WriteLine(result);
// 如果有可选参数
var options = new Dictionary <string, object> {
{ "face_field", "age,beauty,expression,faceshape,gender,glasses,landmark,race,quality,facetype" },
{ "max_face_num", 2 },
{ "face_type", "LIVE" }
};
// 带参数调用人脸检测
var result = client.Detect(image, imageType, options);
Console.WriteLine(result);
}
public JObject FaceJianCe(Stream faceStream)
{
byte[] buffer = new byte[faceStream.Length];
faceStream.Read(buffer, 0, (int)faceStream.Length);
// 调用人脸检测，可能会抛出网络等异常，请使用try/catch捕获
// 如果有可选参数
var options = new Dictionary <string, object> {
{ "max_face_num", 1 },
{ "face_fields", "beauty" }
};
// 带参数调用人脸检测
try
{
return(Client.Detect(buffer, options));
}
catch (Exception ex)
{
throw ex;
}
}
/// <summary>
/// 人脸检测
/// </summary>
/// <param name="sender"></param>
/// <param name="e"></param>
private void skinButton1_Click(object sender, EventArgs e)
{
if (picbPreview.Image is null)
{
MessageBox.Show("请先拍照或者选取一张图片,再进行检测操作");
return;
}
// var APP_ID = "17894506";
Thread threadadd = new Thread(() =>
{
try
{
var client = new Baidu.Aip.Face.Face(API_KEY, SECRET_KEY)
{
Timeout = 60000 // 修改超时时间
};
var image = ImgToBase64String((Bitmap)this.picbPreview.Image);
var imageType = "BASE64";
// 调用人脸检测，可能会抛出网络等异常，请使用try/catch捕获
var result = client.Detect(image, imageType);
// 如果有可选参数
var options = new Dictionary <string, object>
{
{ "face_field", "age,beauty,expression,face_shape,gender,glasses," +
",race,quality,eye_status,emotion,face_type,eye_status" },
{ "max_face_num", Max_face_num },
{ "face_type", "LIVE" },
{ "liveness_control", "LOW" }
};
// 带参数调用人脸检测
result = client.Detect(image, imageType, options);
if (result != null && result.ToString() != null && result.ToString().Length > 0)
{
var json = JsonConvert.SerializeObject(result);
FaceCheckModel model = DeserializeJsonToObject <GetApiJson <FaceCheckModel> >(json)?.Result ?? new FaceCheckModel();
FaceTestBind(model);
}
;
var picclient = new Baidu.Aip.ImageClassify.ImageClassify("NpBGfUR6qBGtFo5bIFbiPCO9", "S0L7LXAewfW7BBKmbXd0EQ8iRzEYRGqc")
{
Timeout = 60000 // 修改超时时间
};
Image img = this.picbPreview.Image;
MemoryStream ms = new MemoryStream();
byte[] imagedata = null;
img.Save(ms, System.Drawing.Imaging.ImageFormat.Jpeg);
imagedata = ms.GetBuffer();
var picoptions = new Dictionary <string, object> {
{ "baike_num", Result_num }
};
var results = picclient.AdvancedGeneral(imagedata, picoptions);
if (results != null && results.ToString() != null && results.ToString().Length > 0)
{
var json = JsonConvert.SerializeObject(results);
ImageRecognitionModel model = DeserializeJsonToObject <ImageRecognitionModel>(json) ?? new ImageRecognitionModel();
ImageRecognitionBind(model);
}
}
catch (Exception ex)
{
MessageBox.Show("网络错误！错误信息:" + ex.Message);
}
});
threadadd.Start();
}
/// <summary>
/// 人脸识别
/// </summary>
/// <returns></returns>
public JsonResult FaceDistinguish()
{
// 设置APPID/AK/SK
var API_KEY = "mxzzu1vLxca9KjnLwBCgOZs5"; //你的 Api Key
var SECRET_KEY = "D9CkVbdziW9GrHiAZDENt8rOf0tVw9im"; //你的 Secret Key
var client = new Baidu.Aip.Face.Face(API_KEY, SECRET_KEY);
client.Timeout = 60000; // 修改超时时间
var imageType = "BASE64"; //BASE64 URL
string imgData64 = Request["imgData64"];
imgData64 = imgData64.Substring(imgData64.IndexOf(",") + 1); //将‘，’以前的多余字符串删除
ResultInfo result = new ResultInfo();
try
{
var groupId = "TestGroupA";
var userId = "TestUserA";
var result323 = client.Detect(imgData64, imageType);
//活体检测阈值是多少
//0.05 活体误拒率：万分之一；拒绝率：63.9%
//0.3 活体误拒率：千分之一；拒绝率：90.3%
//0.9 活体误拒率：百分之一；拒绝率：97.6%
//1误拒率: 把真人识别为假人的概率. 阈值越高，安全性越高, 要求也就越高, 对应的误识率就越高
//2、通过率=1-误拒率
//所以你thresholds参数返回 和 face_liveness 比较大于推荐值就是活体
////活体判断
var faces = new JArray
{
new JObject
{
{ "image", imgData64 },
{ "image_type", "BASE64" }
}
};
var Living = client.Faceverify(faces); //活体检测交互返回
var LivingJson = Newtonsoft.Json.JsonConvert.SerializeObject(Living);
var LivingObj = Newtonsoft.Json.JsonConvert.DeserializeObject(LivingJson) as JObject;
if (LivingObj["error_code"].ToString() == "0" && LivingObj["error_msg"].ToString() == "SUCCESS")
{
var Living_result = Newtonsoft.Json.JsonConvert.DeserializeObject(LivingObj["result"].ToString()) as JObject;
var Living_list = Living_result["thresholds"];
double face_liveness = Convert.ToDouble(Living_result["face_liveness"]);
var frr = Newtonsoft.Json.JsonConvert.SerializeObject(Living_list.ToString());
var frr_1eObj = Newtonsoft.Json.JsonConvert.DeserializeObject(Living_list.ToString()) as JObject;
double frr_1e4 = Convert.ToDouble(frr_1eObj["frr_1e-4"]);
if (face_liveness < frr_1e4)
{
result.info = "识别失败：这是相片之类的非活体！";
return(Json(result, JsonRequestBehavior.AllowGet));
}
}
//首先查询是否存在人脸
var result2 = client.Search(imgData64, imageType, groupId);
var strJson = Newtonsoft.Json.JsonConvert.SerializeObject(result2);
var o2 = Newtonsoft.Json.JsonConvert.DeserializeObject(strJson) as JObject;
//判断是否存在当前人脸，相识度是否大于80
if (o2["error_code"].ToString() == "0" && o2["error_msg"].ToString() == "SUCCESS")
{
var result_list = Newtonsoft.Json.JsonConvert.DeserializeObject(o2["result"].ToString()) as JObject;
var user_list = result_list["user_list"];
var Obj = JArray.Parse(user_list.ToString());
foreach (var item in Obj)
{
//80分以上可以判断为同一人，此分值对应万分之一误识率
var score = Convert.ToInt32(item["score"]);
if (score > 80)
{
result.info = result2.ToString();
result.res = true;
result.startcode = 221;
return(Json(result, JsonRequestBehavior.AllowGet));
}
}
}
else
{
result.info = strJson.ToString();
result.res = false;
return(Json(result, JsonRequestBehavior.AllowGet));
}
}
catch (Exception e)
{
result.info = e.Message;
}
return(Json(result, JsonRequestBehavior.AllowGet));
}
private void button_recongnition2_Click(object sender, RoutedEventArgs e)
{
try
{
OpenFileDialog fdlg = new OpenFileDialog();
fdlg.Title = "选择识别语音文件";
//fdlg.InitialDirectory = @"c:\";
fdlg.Filter = "图片文件(*.jpg,*.jpeg,*.png,*.bmp)|*.jpg;*.jpeg;*.png;*.bmp";
fdlg.FilterIndex = 0;
fdlg.RestoreDirectory = false;
if (fdlg.ShowDialog() == true)
{
m_imgPath2 = fdlg.FileName;
}
string image = Convert.ToBase64String(File.ReadAllBytes(m_imgPath2));
// 调用人脸检测，可能会抛出网络等异常，请使用try/catch捕获
// 如果有可选参数
var options = new Dictionary <string, object>
{
{ "face_field", "beauty,age,expression,face_shape,gender,race" },
};
// 带参数调用人脸检测
var result = mClient.Detect(image, "BASE64", options);
var list_faceinfo = new List <string>();
var faces = result["result"]["face_list"];
foreach (var face in faces)
{
var age = (int)face["age"];
list_faceinfo.Add("年龄：" + age);
var beauty = (string)face["beauty"];
list_faceinfo.Add("样貌评分：" + beauty);
var expression = (string)face["expression"]["type"];
var expression_pb = (string)face["expression"]["probability"];
list_faceinfo.Add("表情：" + expression + ",置信度：" + expression_pb);
var face_shape = (string)face["face_shape"]["type"];
var face_shape_pb = (string)face["face_shape"]["probability"];
list_faceinfo.Add("脸型：" + face_shape + ",置信度：" + face_shape_pb);
var gender = (string)face["gender"]["type"];
var gender_pb = (string)face["gender"]["probability"];
list_faceinfo.Add("性别：" + gender + ",置信度：" + gender_pb);
var race = (string)face["race"]["type"];
var race_pb = (string)face["race"]["probability"];
list_faceinfo.Add("种族：" + race + ",置信度：" + race_pb);
var left = (int)face["location"]["left"];
var top = (int)face["location"]["top"];
var width = (int)face["location"]["width"];
var height = (int)face["location"]["height"];
//var rotation = (int) face["location"]["rotation"];
Rectangle rect = new Rectangle(left, top, width, height);
Image change_img = Image.FromFile(m_imgPath2);
Graphics plate = Graphics.FromImage(change_img);
plate.DrawRectangle(new System.Drawing.Pen(Color.BlueViolet, 2), rect);
plate.Save();
plate.Dispose();
MemoryStream ms = new MemoryStream();
change_img.Save(ms, ImageFormat.Bmp);
change_img.Dispose();
BitmapImage bi = new BitmapImage();
bi.BeginInit();
bi.StreamSource = ms;
bi.EndInit();
image_recongnition2.Source = bi;
image_recongnition2.InvalidateVisual();
break;
}
textBox_recongnition2.Text = string.Join("\n", list_faceinfo);
}
catch (Exception exception)
{
MessageBox.Show(exception.Message, "错误");
}
}
