41

GitHub - xikuqi/OpenCV

 4 years ago
source link: https://github.com/xikuqi/OpenCV
Go to the source link to view the article. You can view the picture content, updated content and better typesetting reading experience. If the link is broken, please click the button below to view the snapshot at that time.

README.md

seetafaceJNI

项目介绍

基于中科院seetaface2进行封装的JAVA人脸识别算法库,支持人脸识别、1:1比对、1:N比对。 seetaface2:https://github.com/seetaface/SeetaFaceEngine2

环境配置

1、下载model( https://pan.baidu.com/s/1HJj8PEnv3SOu6ZxVpAHPXg ) 文件到本地,并解压出来;

2、下载doc目录中对应的lib包到本地并解压:Windows(64位)环境下载lib-win-x64.zip、Linux(64位)下载lib-linux-x64.tar.bz2,Linux环境还需要安装依赖库,详见:https://my.oschina.net/u/1580184/blog/3042404

3、将doc中的faces-data.db下载到本地;(PS:如果不需要使用1:N人脸搜索,不需要此文件,需要将seetafce.properties中的sqlite.db.file配置注释掉);

4、将src/main/resources/中的seetaface.properties文件放到项目的resources根目录中;

#linux系统中依赖的lib名称
libs=holiday,SeetaFaceDetector200,SeetaPointDetector200,SeetaFaceRecognizer200,SeetaFaceCropper200,SeetaFace2JNI
#Windows系统中依赖的lib名称
#libs=libgcc_s_sjlj-1,libeay32,libquadmath-0,ssleay32,libgfortran-3,libopenblas,holiday,SeetaFaceDetector200,SeetaPointDetector200,SeetaFaceRecognizer200,SeetaFaceCropper200,SeetaFace2JNI

#lib存放目录
libs.path=/usr/local/seetaface2/lib
#model存放目录
bindata.dir=/usr/local/seetaface2/bindata

##sqlite配置(如果不用1:N人脸搜索功能,请删除下面5项sqlite开头的配置)
sqlite.db.file=/data/faces-data.db
sqlite.conn.maxTotal=50
sqlite.conn.maxIdle=5
sqlite.conn.minIdle=0
sqlite.conn.maxWaitMillis=60000

5、将seetafaceJNI-2.0.jar和依赖包导入到项目中,pom如下:

   <properties>
       <spring.version>4.2.8.RELEASE</spring.version>
       <log4j.version>2.8.2</log4j.version>
       <slf4j.version>1.7.25</slf4j.version>
   </properties>
  
   <dependencies>
       <dependency>
            <groupId>com.cnsugar.ai</groupId>
            <artifactId>seetafaceJNI</artifactId>
            <version>2.0</version>
            <!--<scope>system</scope>-->
            <!--<systemPath>${project.basedir}/lib/seetafaceJNI-2.0.jar</systemPath>-->
       </dependency>
       <dependency>
           <groupId>org.springframework</groupId>
           <artifactId>spring-core</artifactId>
           <version>${spring.version}</version>
       </dependency>
  
       <dependency>
           <groupId>org.slf4j</groupId>
           <artifactId>slf4j-api</artifactId>
           <version>${slf4j.version}</version>
       </dependency>
  
       <!-- sqlite -->
       <dependency>
           <groupId>org.xerial</groupId>
           <artifactId>sqlite-jdbc</artifactId>
           <version>3.25.2</version>
       </dependency>
       <dependency>
           <groupId>org.apache.commons</groupId>
           <artifactId>commons-pool2</artifactId>
           <version>2.4.2</version>
       </dependency>
   </dependencies> 

6、调用FaceHelper中的方法。

使用方法

所有方法都封装到了FaceHelper工具类中

    /**
     * 人脸比对
     *
     * @param img1
     * @param img2
     * @return 相似度
     */
    float compare(File img1, File img2);
    float compare(byte[] img1, byte[] img2);
    float compare(BufferedImage image1, BufferedImage image2);
    
    /**
     * 注册人脸(会裁剪图片)
     *
     * @param key 人脸照片唯一标识
     * @param img 人脸照片
     * @return 
     */
    boolean register(String key, byte[] img);
    /**
     * 注册人脸(不裁剪图片)
     *
     * @param key 人脸照片唯一标识
     * @param image 人脸照片
     * @return 
     */
    boolean register(String key, BufferedImage image)
    
    /**
     * 搜索人脸
     *
     * @param img 人脸照片
     * @return
     */
    Result search(byte[] img);
    Result search(BufferedImage image);
    
    /**
     * 人脸提取(裁剪)
     *
     * @param img
     * @return return cropped face
     */
    BufferedImage crop(byte[] img);
    BufferedImage crop(BufferedImage image);
    
    /**
     * 人脸识别
     *
     * @param img
     * @return
     */
    SeetaRect[] detect(byte[] img);
    SeetaRect[] detect(BufferedImage image);

    /**
     * 人脸识别(包含5个特征点位置)
     *
     * @param image
     * @return
     */
    FaceLandmark detectLandmark(BufferedImage image);
    
    /**
     * 删除已注册的人脸
     * @param keys
     */
    void removeRegister(String... keys);  
    
    /**
     * 清除人脸库数据
     */
    void clear();    
    
  • 示例代码:1:1人脸比对
    @org.junit.Test
    public void testCompare() throws Exception {
        String img1 = "F:\\ai\\demo-pic39.jpg";
        String img2 = "F:\\ai\\left_pic_one.jpg";
        System.out.println("result:"+FaceHelper.compare(new File(img1), new File(img2)));
    }
  • 示例代码:1:N人脸搜索 先调用FaceHelper.register()方法将人脸图片注册到seetaface2的人脸库(内存)中,同时会将图片存在sqlite数据库中进行持久化,下次应用程序启动时会自动从sqlite中把图片读取出来重新注册到seetafce2的内存库中
    @org.junit.Test
    public void testRegister() throws IOException {
        //将F:\ai\star目录下的jpg、png图片都注册到人脸库中,以文件名为key
        Collection<File> files = FileUtils.listFiles(new File("F:\\ai\\star"), new String[]{"jpg", "png"}, false);
        for (File file : files) {
            String key = file.getName();
            try {
                FaceHelper.register(key, FileUtils.readFileToByteArray(file));
            } catch (Exception e) {
                e.printStackTrace();
            }
        }
    }

    @org.junit.Test
    public void testSearch() throws IOException {
        SeetafaceBuilder.build();//系统启动时先调用初始化方法

        //等待初始化完成
        while (SeetafaceBuilder.getFaceDbStatus() == SeetafaceBuilder.FacedbStatus.LOADING || SeetafaceBuilder.getFaceDbStatus() == SeetafaceBuilder.FacedbStatus.READY) {
            try {
                Thread.sleep(100);
            } catch (InterruptedException e) {
                e.printStackTrace();
            }
        }
        
        long l = System.currentTimeMillis();
        Result result = FaceHelper.search(FileUtils.readFileToByteArray(new File("F:\\ai\\gtl.jpg")));
        System.out.println("搜索结果:" + result + ", 耗时:" + (System.currentTimeMillis() - l));
    }

About Joyk


Aggregate valuable and interesting links.
Joyk means Joy of geeK