当前位置: 首页>>代码示例>>C#>>正文


C# Face.SetRecognitionData方法代码示例

本文整理汇总了C#中Face.SetRecognitionData方法的典型用法代码示例。如果您正苦于以下问题:C# Face.SetRecognitionData方法的具体用法?C# Face.SetRecognitionData怎么用?C# Face.SetRecognitionData使用的例子?那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在Face的用法示例。


在下文中一共展示了Face.SetRecognitionData方法的1个代码示例,这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞,您的评价将有助于系统推荐出更棒的C#代码示例。

示例1: RecognizeMultipleFaces

        /// <summary>
        /// Recognizes multiple faces from a single image frame
        /// </summary>
        /// <param name="image">Neurotec Image in which is based the face recognition</param>
        /// <param name="vlFaces">Array of faces detected</param>
        /// <param name="recognizedFaces">Array of Face objects used to realize each recognition</param>
        /// <param name="MultipleRecognitionResults">An array containing all recognition results for each recognized face</param>
        /// <returns>An array containing best match in all known faces.</returns>
        private RecognitionResult[] RecognizeMultipleFaces(NImage image, VleFace[] vlFaces, out FaceCollection detectedFaces, out RecognitionResult[][] MultipleRecognitionResults)
        {
            #region Variables
            // Stores the original image as bitmap
            Bitmap bmp;
            // Bitmap to draw in the detected face region
            Bitmap croppedBitmap;
            // Graphics used to copy the face detected region
            Graphics g;
            // Rectangle used to copy the scaled region of face detected
            Rectangle rect;
            // Nurotec Image required in the process of recognize the face detected region
            NGrayscaleImage gray;
            // Verilook Detetion Details as result of face recognition
            VleDetectionDetails detectionDetails;
            // The face template result of a face recognition
            byte[][] templates = new byte[vlFaces.Length][];
            // The face features result of a face recognition
            byte[] features;
            // Stores the current recognition face
            Face currentFace;
            // Stores the recognized faces
            //FaceCollection recognizedFaces = new FaceCollection(vlFaces.Length);
            detectedFaces = new FaceCollection(vlFaces.Length);
            // Stores the best recognition result for current face
            RecognitionResult currentResult;
            // Stores the recognition results for current face
            RecognitionResult[] currentRecognitionResults;
            // Stores all Recognition results
            List<RecognitionResult[]> recognitionResults = new List<RecognitionResult[]>();
            // Stores the best recognition result matches
            List<RecognitionResult> selectedResults = new List<RecognitionResult>();
            #endregion

            // Get the original image as bitmap
            bmp = new Bitmap(image.ToBitmap());

            // Extract each face, and get its template
            foreach (VleFace vlFace in vlFaces)
            {
                // Get a rectangle a bit larger than the one the face has been recognized.
                // Its because some times in the exact area of the face the face cannot be recognized again
                //rect = new Rectangle(vlFace.Rectangle.X - 50, vlFace.Rectangle.Y - 50, vlFace.Rectangle.Width + 100, vlFace.Rectangle.Height + 100);
                rect = new Rectangle(vlFace.Rectangle.X - vlFace.Rectangle.Width / 2, vlFace.Rectangle.Y - vlFace.Rectangle.Height / 2, vlFace.Rectangle.Width * 2, vlFace.Rectangle.Height * 2);
                // Get the face bitmap
                croppedBitmap = new Bitmap(rect.Width, rect.Height);
                g = Graphics.FromImage(croppedBitmap);
                g.DrawImage(bmp, 0, 0, rect, GraphicsUnit.Pixel);
                // Get gray image for face detection
                gray = (NGrayscaleImage)NImage.FromImage(NPixelFormat.Grayscale, 0, NImage.FromBitmap(croppedBitmap));

                // Extract the face and extract its template
                currentFace = new Face(vlFace);
                features = vlExtractor.Extract(gray, out detectionDetails);
                if (!detectionDetails.FaceAvailable) continue;
                UseResources();
                currentFace.SetRecognitionData(features, detectionDetails, croppedBitmap);
                ReleaseResources();
                currentFace.CalculateFovAndCoords((int)image.Width, (int)image.Height);
                detectedFaces.Add(currentFace);
                Console("Found face: location = (" + detectionDetails.Face.Rectangle.X + ", " + detectionDetails.Face.Rectangle.Y + "), width = " + detectionDetails.Face.Rectangle.Width + ", height = " + detectionDetails.Face.Rectangle.Height + ", confidence = " + detectionDetails.Face.Confidence);

                try
                {
                    croppedBitmap.Dispose();
                    g.Dispose();
                    gray.Dispose();
                }
                catch { }

            }
            if (detectedFaces.Count > 0) Console(detectedFaces.Count.ToString() + " faces found.");
            if (knownFaces.Count > 0)
            {
                Console("Initializing recognition");
                // Recognize each detected face
                for (int i = 0; i < detectedFaces.Count; ++i)
                {
                    if (detectedFaces[i].Features == null) continue;
                    currentFace = detectedFaces[i];

                    // Start recognition
                    currentResult = Recognize(currentFace, out currentRecognitionResults);
                    if (currentResult == null) continue;
                    selectedResults.Add(currentResult);
                    recognitionResults.Add(currentRecognitionResults);
                }
            }

            MultipleRecognitionResults = recognitionResults.ToArray();
            return selectedResults.ToArray();
        }
开发者ID:BioRoboticsUNAM,项目名称:PRS-FND,代码行数:100,代码来源:HumanRecognizer.cs


注:本文中的Face.SetRecognitionData方法示例由纯净天空整理自Github/MSDocs等开源代码及文档管理平台,相关代码片段筛选自各路编程大神贡献的开源项目,源码版权归原作者所有,传播和使用请参考对应项目的License;未经允许,请勿转载。