修复人脸识别算法facenet-pytorch实现方式

examples/pom.xml

@@ -43,7 +43,7 @@
         <dependency>
             <groupId>ink.numberone</groupId>
             <artifactId>smartjavaai-face</artifactId>
-            <version>1.0.5</version>
+            <version>1.0.6</version>
         </dependency>
 
         <dependency>

examples/src/main/java/smartai/examples/face/FaceDemo.java

@@ -13,8 +13,10 @@ import smartai.examples.utils.ImageUtils;
 import javax.imageio.ImageIO;
 import java.awt.*;
 import java.awt.image.BufferedImage;
+import java.awt.image.RasterFormatException;
 import java.io.File;
 import java.io.FileInputStream;
+import java.io.IOException;
 import java.net.URL;
 import java.nio.file.Files;
 import java.nio.file.LinkOption;
@@ -30,7 +32,7 @@ public class FaceDemo {
 
     public static void main(String[] args) {
         try {
-            featureComparison();
+            featureExtractionAndCompare2();
             //detectFace2();
             //verifyIDCard();
         } catch (Exception e) {
@@ -44,29 +46,26 @@ public class FaceDemo {
      * 特点：识别精度高，高速
      * 应用场景：如监控摄像头、智能安防系统等需要高精度检测的场合
      */
-    public static void detectFace() throws Exception {
-        // 创建并启动计时器
-        StopWatch sw = StopWatch.createStarted();
-        //创建人脸算法
-        FaceAlgorithm currentAlgorithm = FaceAlgorithmFactory.createFaceAlgorithm();
-        sw.stop();
-        log.info("创建人脸算法耗时：" + sw.getTime() + "ms");
-        sw.reset();
-        sw.start();
-        //使用图片路径检测
-        FaceDetectedResult result = currentAlgorithm.detect("src/main/resources/largest_selfie.jpg");
-        sw.stop();
-        log.info("人脸检测耗时：" + sw.getTime() + "ms");
-        log.info("人脸检测结果：{}", JSONObject.toJSONString(result));
-        //使用图片流检测
-        File input = new File("src/main/resources/largest_selfie.jpg");
-        //FaceDetectedResult result = currentAlgorithm.detect(new FileInputStream(input));
-        //log.info("人脸检测结果：{}", JSONObject.toJSONString(result));
-        BufferedImage image = ImageIO.read(input);
-        //创建保存路径
-        Path imagePath = Paths.get("output").resolve("retinaface_detected.jpg");
-        //绘制人脸框
-        ImageUtils.drawBoundingBoxes(image, result, imagePath.toAbsolutePath().toString());
+    public static void detectFace(){
+        try {
+            //创建人脸算法
+            FaceAlgorithm currentAlgorithm = FaceAlgorithmFactory.createFaceAlgorithm();
+            //使用图片路径检测
+            FaceDetectedResult result = currentAlgorithm.detect("src/main/resources/largest_selfie.jpg");
+            log.info("人脸检测结果：{}", JSONObject.toJSONString(result));
+            //使用图片流检测
+            File input = new File("src/main/resources/largest_selfie.jpg");
+            //FaceDetectedResult result = currentAlgorithm.detect(new FileInputStream(input));
+            //log.info("人脸检测结果：{}", JSONObject.toJSONString(result));
+            BufferedImage image = ImageIO.read(input);
+            //创建保存路径
+            Path imagePath = Paths.get("output").resolve("retinaface_detected.jpg");
+            //绘制人脸框
+            ImageUtils.drawBoundingBoxes(image, result, imagePath.toAbsolutePath().toString());
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+
     }
 
 
@@ -76,29 +75,25 @@ public class FaceDemo {
      * 特点：高速，准确率略低
      * 应用场景：如监控摄像头、智能安防系统等需要高精度检测的场合
      */
-    public static void detectFace2() throws Exception {
-        // 创建并启动计时器
-        StopWatch sw = StopWatch.createStarted();
-        //创建轻量人脸算法
-        FaceAlgorithm currentAlgorithm = FaceAlgorithmFactory.createLightFaceAlgorithm();
-        sw.stop();
-        log.info("创建人脸算法耗时：" + sw.getTime() + "ms");
-        sw.reset();
-        sw.start();
-        //使用图片路径检测
-        FaceDetectedResult result = currentAlgorithm.detect("src/main/resources/largest_selfie.jpg");
-        sw.stop();
-        log.info("人脸检测耗时：" + sw.getTime() + "ms");
-        log.info("轻量人脸检测结果：{}", JSONObject.toJSONString(result));
-        //使用图片流检测
-        //File imageFile = new File("/Users/wenjie/Downloads/djl-master/examples/src/test/resources/largest_selfie.jpg");
-        //FaceDetectedResult result = currentAlgorithm.detect(new FileInputStream(imageFile));
-        File input = new File("src/main/resources/largest_selfie.jpg");
-        BufferedImage image = ImageIO.read(input);
-        //创建保存路径
-        Path imagePath = Paths.get("output").resolve("retinaface_detected.jpg");
-        //绘制人脸框
-        ImageUtils.drawBoundingBoxes(image, result, imagePath.toAbsolutePath().toString());
+    public static void detectFace2(){
+        try {
+            //创建轻量人脸算法
+            FaceAlgorithm currentAlgorithm = FaceAlgorithmFactory.createLightFaceAlgorithm();
+            //使用图片路径检测
+            FaceDetectedResult result = currentAlgorithm.detect("src/main/resources/largest_selfie.jpg");
+            log.info("轻量人脸检测结果：{}", JSONObject.toJSONString(result));
+            //使用图片流检测
+            //File imageFile = new File("/Users/wenjie/Downloads/djl-master/examples/src/test/resources/largest_selfie.jpg");
+            //FaceDetectedResult result = currentAlgorithm.detect(new FileInputStream(imageFile));
+            File input = new File("src/main/resources/largest_selfie.jpg");
+            BufferedImage image = ImageIO.read(input);
+            //创建保存路径
+            Path imagePath = Paths.get("output").resolve("retinaface_detected.jpg");
+            //绘制人脸框
+            ImageUtils.drawBoundingBoxes(image, result, imagePath.toAbsolutePath().toString());
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
     }
 
 
@@ -109,34 +104,38 @@ public class FaceDemo {
      * 特点：识别精度高，高速
      * 应用场景：如监控摄像头、智能安防系统等需要高精度检测的场合
      */
-    public static void detectFaceOffine() throws Exception {
-        // 初始化配置
-        ModelConfig config = new ModelConfig();
-        config.setAlgorithmName("retinaface");//人脸算法模型，目前支持：retinaface/ultralightfastgenericface/seetaface6
-        //config.setAlgorithmName("ultralightfastgenericface");//轻量模型
-        config.setConfidenceThreshold(FaceConfig.DEFAULT_CONFIDENCE_THRESHOLD);//置信度阈值
-        config.setMaxFaceCount(FaceConfig.MAX_FACE_LIMIT);//每张特征图保留的最大候选框数量
-        //nms阈值:控制重叠框的合并程度,取值越低，合并越多重叠框（减少误检但可能漏检）；取值越高，保留更多框（增加检出但可能引入冗余）
-        config.setNmsThresh(FaceConfig.NMS_THRESHOLD);
-        //模型下载地址：
-        //retinaface: https://resources.djl.ai/test-models/pytorch/retinaface.zip
-        //ultralightfastgenericface: https://resources.djl.ai/test-models/pytorch/ultranet.zip
-        //改为模型存放路径
-        config.setModelPath("/Users/wenjie/Documents/develop/face_model/retinaface.pt");
-        //创建人脸算法
-        FaceAlgorithm currentAlgorithm = FaceAlgorithmFactory.createFaceAlgorithm(config);
-        //使用图片路径检测
-        FaceDetectedResult result = currentAlgorithm.detect("src/main/resources/largest_selfie.jpg");
-        log.info("人脸检测结果：{}", JSONObject.toJSONString(result));
-        //使用图片流检测
-        File input = new File("src/main/resources/largest_selfie.jpg");
-        //FaceDetectedResult result = currentAlgorithm.detect(new FileInputStream(input));
-        //logger.info("人脸检测结果：{}", JSONObject.toJSONString(result));
-        BufferedImage image = ImageIO.read(input);
-        //创建保存路径
-        Path imagePath = Paths.get("output").resolve("retinaface_detected.jpg");
-        //绘制人脸框
-        ImageUtils.drawBoundingBoxes(image, result, imagePath.toAbsolutePath().toString());
+    public static void detectFaceOffine(){
+        try {
+            // 初始化配置
+            ModelConfig config = new ModelConfig();
+            config.setAlgorithmName("retinaface");//人脸算法模型，目前支持：retinaface/ultralightfastgenericface/seetaface6
+            //config.setAlgorithmName("ultralightfastgenericface");//轻量模型
+            config.setConfidenceThreshold(FaceConfig.DEFAULT_CONFIDENCE_THRESHOLD);//置信度阈值
+            config.setMaxFaceCount(FaceConfig.MAX_FACE_LIMIT);//每张特征图保留的最大候选框数量
+            //nms阈值:控制重叠框的合并程度,取值越低，合并越多重叠框（减少误检但可能漏检）；取值越高，保留更多框（增加检出但可能引入冗余）
+            config.setNmsThresh(FaceConfig.NMS_THRESHOLD);
+            //模型下载地址：
+            //retinaface: https://resources.djl.ai/test-models/pytorch/retinaface.zip
+            //ultralightfastgenericface: https://resources.djl.ai/test-models/pytorch/ultranet.zip
+            //改为模型存放路径
+            config.setModelPath("/Users/wenjie/Documents/develop/face_model/retinaface.pt");
+            //创建人脸算法
+            FaceAlgorithm currentAlgorithm = FaceAlgorithmFactory.createFaceAlgorithm(config);
+            //使用图片路径检测
+            FaceDetectedResult result = currentAlgorithm.detect("src/main/resources/largest_selfie.jpg");
+            log.info("人脸检测结果：{}", JSONObject.toJSONString(result));
+            //使用图片流检测
+            File input = new File("src/main/resources/largest_selfie.jpg");
+            //FaceDetectedResult result = currentAlgorithm.detect(new FileInputStream(input));
+            //logger.info("人脸检测结果：{}", JSONObject.toJSONString(result));
+            BufferedImage image = ImageIO.read(input);
+            //创建保存路径
+            Path imagePath = Paths.get("output").resolve("retinaface_detected.jpg");
+            //绘制人脸框
+            ImageUtils.drawBoundingBoxes(image, result, imagePath.toAbsolutePath().toString());
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
     }
 
     /**
@@ -164,7 +163,8 @@ public class FaceDemo {
     }
 
     /**
-     * 人脸特征提取及比对
+     * seetaface6人脸特征提取及比对（可人证核验）
+     * 目前仅支持windows 64位系统，如需支持其他操作系统可参考方法：featureExtractionAndCompare2
      */
     public static void featureExtractionAndCompare(){
         try {
@@ -180,8 +180,12 @@ public class FaceDemo {
             //提取图像中最大人脸的特征
             float[] feature1 = currentAlgorithm.featureExtraction("src/main/resources/kana1.jpg");
             float[] feature2 = currentAlgorithm.featureExtraction("src/main/resources/kana2.jpg");
-            float similar = currentAlgorithm.calculSimilar(feature1, feature2);
-            log.info("相似度：{}", similar);
+            if(feature1 != null && feature2 != null){
+                float similar = currentAlgorithm.calculSimilar(feature1, feature2);
+                log.info("相似度：{}", similar);
+            }else{
+                log.warn("人脸特征提取失败");
+            }
         }
         catch (Exception e){
             e.printStackTrace();
@@ -189,8 +193,31 @@ public class FaceDemo {
     }
 
     /**
-     * 注册人脸及搜索人脸（1：N）
+     * facenet-pytorch 人脸特征提取及比对（可人证核验）
+     * 支持windows，linux，macos
      */
+    public static void featureExtractionAndCompare2(){
+        try {
+            //创建脸算法
+            FaceAlgorithm featureAlgorithm = FaceAlgorithmFactory.createFaceFeatureAlgorithm();
+            //提取身份证人脸特征
+            float[] feature1 = featureAlgorithm.featureExtraction("src/main/resources/kana1.jpg");
+            float[] feature2 = featureAlgorithm.featureExtraction("src/main/resources/kana2.jpg");
+            if (feature1 != null && feature2 != null) {
+                //相似度在0.8至0.85及以上时，可判定为同一人，但具体阈值可能因图片而异，存在一定误差。
+                float similar = featureAlgorithm.calculSimilar(feature1, feature2);
+                log.info("相似度：{}", similar);
+            } else {
+                log.warn("人脸特征提取失败");
+            }
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+        /**
+         * 注册人脸及搜索人脸（1：N）
+         */
     public static void registerAndSearchFace(){
         try {
             // 初始化配置
@@ -248,5 +275,4 @@ public class FaceDemo {
         }
     }
 
-
 }