opencv - 从 hbase 读取图像并使用 Opencv 检测该图像中的人脸

Question

我需要读取 hbase 中的图像并转换为 opencv mat 以进行人脸检测。
我的代码如下

public static class FaceCountMapper extends TableMapper<Text, Text> {
private CascadeClassifier faceDetector;

public void setup(Context context) throws IOException, InterruptedException {

    if (context.getCacheFiles() != null && context.getCacheFiles().length > 0) {
        URI mappingFileUri = context.getCacheFiles()[0];

        if (mappingFileUri != null) {
            System.out.println(mappingFileUri);
            faceDetector = new CascadeClassifier(mappingFileUri.toString());

        }
    }

    super.setup(context);
} // setup()

public ArrayList<Object> detectFaces(Mat image, String file_name) {

    ArrayList<Object> facemap = new ArrayList<Object>();

    MatOfRect faceDetections = new MatOfRect();

    faceDetector.detectMultiScale(image, faceDetections);

    System.out.println(String.format("Detected %s faces", faceDetections.toArray().length));
    output.put(faceDetections.toArray().length);

    facemap.add(output);

}

return facemap;

}

public void map(ImmutableBytesWritable row, Result result, Context context)
        throws InterruptedException, IOException {

    String file_name = Bytes.toString(result.getValue(Bytes.toBytes("Filename"), Bytes.toBytes("data")));

    String mimetype = Bytes.toString(result.getValue(Bytes.toBytes("mime"), Bytes.toBytes("data")));

    byte[] image_data = result.getValue(Bytes.toBytes("Data"), Bytes.toBytes("data"));

    BufferedImage bi = ImageIO.read(new ByteArrayInputStream(image_data));

    Mat mat = new Mat(bi.getHeight(), bi.getWidth(), CvType.CV_8UC3);

    mat.put(0, 0, image_data);
    detectFaces(mat, file_name);

}

作业配置如下

Configuration conf = this.getConf();
    conf.set("hbase.master", "101.192.0.122:16000");
    conf.set("hbase.zookeeper.quorum", "101.192.0.122");
    conf.setInt("hbase.zookeeper.property.clientPort", 2181);
    conf.set("zookeeper.znode.parent", "/hbase-unsecure");

    // Initialize and configure MapReduce job
    Job job = Job.getInstance(conf);

    job.setJarByClass(FaceCount3.class);
    job.setMapperClass(FaceCountMapper.class);
    job.getConfiguration().set("fs.hdfs.impl", org.apache.hadoop.hdfs.DistributedFileSystem.class.getName());
    job.getConfiguration().set("fs.file.impl", org.apache.hadoop.fs.LocalFileSystem.class.getName());

    Scan scan = new Scan();
    scan.setCaching(500); // 1 is the default in Scan, which will be bad for
                            // MapReduce jobs
    scan.setCacheBlocks(false); // don't set to true for MR jobs

    TableMapReduceUtil.initTableMapperJob("Image", // input HBase table name
            scan, // Scan instance to control CF and attribute selection
            FaceCountMapper.class, // mapper
            null, // mapper output key
            null, // mapper output value
            job);
    job.setOutputFormatClass(NullOutputFormat.class); // because we aren't
                                                        // emitting anything
                                                        // from mapper

    job.addCacheFile(new URI("/user/hduser/haarcascade_frontalface_alt.xml"));
    job.addFileToClassPath(new Path("/user/hduser/hipi-2.1.0.jar"));
    job.addFileToClassPath(new Path("/user/hduser/javacpp.jar"));
    DistributedCache.addFileToClassPath(new Path("/user/hduser/haarcascade_frontalface_alt.xml"), conf);
    conf.set("mapred.job.tracker", "local");
    // Execute the MapReduce job and block until it complets
    boolean success = job.waitForCompletion(true);

    // Return success or failure
    return success ? 0 : 1;

运行时我得到

java.lang.Exception: java.lang.UnsatisfiedLinkError: org.opencv.objdetect.CascadeClassifier.CascadeClassifier_1(Ljava/lang/String;)J

错误。

但是Opencv.jar在hadoop_classpath中提供

Answer 1

当应用程序尝试加载 native 库时抛出 UnsatisfiedLinkError，例如 Linux 中的 .so、Windows 中的 .dll 或 .dylib在 Mac 中，该库不存在。具体来说，为了找到所需的 native 库，JVM 会同时查看 PATH 环境变量和 java.library.path 系统属性。

此外，如果您的应用程序已经加载了该库，并且该应用程序尝试再次加载它，则 JVM 将抛出 UnsatisfiedLinkError。此外，您必须验证 native 库是否存在于 java.library.path 或应用程序的 PATH 环境库中。如果仍然找不到库，请尝试提供 System.loadLibrary 方法的绝对路径。

在您的情况下，请尝试调用方的以下方法并查看类路径元素是什么。

/**
     * Method printClassPathResources.
     */
    public static void printClassPathResources() {
        final ClassLoader cl = ClassLoader.getSystemClassLoader();
        final URL[] urls = ((URLClassLoader) cl).getURLs();
        LOG.info("Print All Class path resources under currently running class");
        for (final URL url : urls) {
            LOG.info(url.getFile());
        }

    }

根据这些输入，您可以调整您的类路径条目(在本例中为 opencv jar 或其他东西)并查看是否有效。