/// <summary>
/// 获取面部检测
/// </summary>
/// <param name="appId">应用ID</param>
/// <param name="sdkKey">应用Key</param>
/// <param name="age">年龄检测器</param>
/// <param name="gender">性别检测器</param>
/// <param name="useTracking">如果设置为True，则启用面部跟踪，如果False，则启用面部检测</param>
/// <param name="preAllocMemSize">缓存区内存大小（byte）</param>
/// <param name="orientPriority">期望的脸部检测角度范围</param>
/// <param name="scale">
/// 用于数值表示的最小人脸尺寸有效值范围[2,50] 推荐值 16（面部跟踪的取值范围为[2,16]）。该尺寸是人脸相对于所在图片的长边的占比。例如，如果用户想检测到的最小人脸尺寸是图片长度的
/// 1/8，那么这个 nScale 就应该设置为 8
/// </param>
/// <param name="maxFaceNumber">用户期望引擎最多能检测出的人脸数有效值范围[1,50]（面部跟踪的取值范围为[1,20]）</param>
/// <returns></returns>
public static FaceLocator GetLocator(string appId, string sdkKey, FaceExtra age, FaceExtra gender, bool useTracking = true,
int preAllocMemSize = 41943040,
OrientPriority orientPriority = OrientPriority.OrientHigherExt,
int scale = 16, int maxFaceNumber = 10)
{
FaceLocator locator;
if (useTracking)
{
locator = new FaceTracking(appId, sdkKey, age, gender, false, preAllocMemSize);
}
else
{
locator = new FaceDetection(appId, sdkKey, age, gender, false, preAllocMemSize);
}
var code = locator.Initialize(orientPriority, scale, maxFaceNumber);
if (code != ErrorCode.Ok)
{
throw new FaceException(code);
}
return(locator);
}
/// <summary>
/// 初始化人脸检测
/// </summary>
/// <param name="appId">应用ID</param>
/// <param name="sdkKey">应用Key</param>
/// <param name="age">年龄识别模块</param>
/// <param name="gender">性别识别模块</param>
/// <param name="autoInitWithDefaultParameter">如果设置为True，将自动采用默认参数初始化引擎</param>
/// <param name="preAllocMemSize">缓存区内存大小（byte）</param>
public FaceTracking(string appId, string sdkKey, FaceExtra age, FaceExtra gender,
bool autoInitWithDefaultParameter = true,
int preAllocMemSize = 41943040) : base(appId, sdkKey, age, gender, autoInitWithDefaultParameter,
preAllocMemSize)
{
}
/// <summary>
/// 获取面部检测的人脸跟踪，此定位方式速度较快，适合视频类的快速定位，但是精度相对较低(目前该定位器存在问题，官方的DEMO都有问题)
/// </summary>
/// <param name="appId">应用ID</param>
/// <param name="sdkKey">应用Key</param>
/// <param name="age">年龄检测器</param>
/// <param name="gender">性别检测器</param>
/// <param name="preAllocMemSize">缓存区内存大小（byte）</param>
/// <param name="orientPriority">期望的脸部检测角度范围</param>
/// <param name="scale">
/// 用于数值表示的最小人脸尺寸有效值范围[2,16] 推荐值 16。该尺寸是人脸相对于所在图片的长边的占比。例如，如果用户想检测到的最小人脸尺寸是图片长度的 1/8，那么这个 nScale 就应该设置为
/// 8
/// </param>
/// <param name="maxFaceNumber">用户期望引擎最多能检测出的人脸数有效值范围[1,20]</param>
/// <returns></returns>
public static FaceLocator GetTrackingLocator(string appId, string sdkKey, FaceExtra age, FaceExtra gender, int preAllocMemSize = 41943040,
OrientPriority orientPriority = OrientPriority.OrientHigherExt,
int scale = 16, int maxFaceNumber = 10)
{
return(GetLocator(appId, sdkKey, age, gender, true, preAllocMemSize, orientPriority, scale, maxFaceNumber));
}
