point-cloud-library - PCL - 使用 LUM 进行全局注册

Question

我正在做一个注册项目，我有一把椅子，上面有一些物体在 kinect 前旋转。

我可以成功进行成对注册，但正如预期的那样有一些漂移(导致图像)。

我想使用 LUM，以便对累积误差进行全局最小化(然后将其“传播”到帧中)，但最终我让帧漂浮在空中。 (图片下方的代码)

这在LUM的使用上有什么明显的错误吗？
---我使用关键点+特征，而不是盲目地用完整的点云来喂养LUM

为什么所有示例都添加单向边而不是双向边？

PARAM_LUM_centroidDistTHRESH = 0.30;
PARAM_LUM_MaxIterations = 100;
PARAM_LUM_ConvergenceThreshold = 0.0f;
int NeighborhoodNUMB = 2;
int FrameDistLOOPCLOSURE = 5;
PARAM_CORR_REJ_InlierThreshold = 0.020;

pcl::registration::LUM<pcl::PointXYZRGBNormal> lum;
lum.setMaxIterations(        PARAM_LUM_MaxIterations );
lum.setConvergenceThreshold( PARAM_LUM_ConvergenceThreshold );

QVector< pcl::PointCloud<pcl::PointXYZRGB>::Ptr >  cloudVector_ORGan_P_;



for (int iii=0; iii<totalClouds; iii++)
{                
        // read - iii_cloud_ORGan_P_
        // transform it with pairwise registration result
        cloudVector_ORGan_P_.append( iii_cloud_ORGan_P_ );
}

for (size_t iii=0; iii<totalClouds; iii++)
{        
    pcl::compute3DCentroid( *cloudVector_ORGan_P_[iii], centrVector[iii] );

    pcl::IntegralImageNormalEstimation<pcl::PointXYZRGB,pcl::Normal> ne;
    //blah blah parameters
    //compute normals with *ne*                            
    //pcl::removeNaNFromPointCloud
    //pcl::removeNaNNormalsFromPointCloud

    pcl::ISSKeypoint3D< pcl::PointXYZRGBNormal, pcl::PointXYZRGBNormal> keyPointDetector;
    //blah balh parameters;
    //keyPointDetector.compute
    //then remove  NAN keypoints

    pcl::SHOTColorEstimationOMP< pcl::PointXYZRGBNormal,pcl::PointXYZRGBNormal,pcl::SHOT1344 > featureDescriptor;
    //featureDescriptor.setSearchSurface( **ful_unorganized_cloud_in_here** );
    //featureDescriptor.setInputNormals(  **normals_from_above____in_here** );
    //featureDescriptor.setInputCloud(    **keypoints_from_above__in_here** );
    //blah blah parameters
    //featureDescriptor.compute
    //delete NAN *Feature* + corresp. *Keypoints* with *.erase*
}

for (size_t iii=0; iii<totalClouds; iii++)
{
    lum.addPointCloud( KEYptVector_UNorg_P_[iii] );
}

for (size_t iii=1; iii<totalClouds; iii++)
{
    for (size_t jjj=0; jjj<iii; jjj++)
    {
        double cloudCentrDISTANCE = ( centrVector[iii] - centrVector[jjj] ).norm();

            if (   (cloudCentrDISTANCE<PARAM_LUM_centroidDistTHRESH  &&  qAbs(iii-jjj)<=NeighborhoodNUMB)       ||
                   (cloudCentrDISTANCE<PARAM_LUM_centroidDistTHRESH  &&  qAbs(iii-jjj)> FrameDistLOOPCLOSURE)   )
            {

                int sourceID;
                int targetID;

                if (qAbs(iii-jjj)<=NeighborhoodNUMB) // so that connection are e.g. 0->1, 1->2, 2->3, 3->4, 4->5, 5->0
                {                                    // not sure if it helps
                    sourceID = jjj;
                    targetID = iii;
                }
                else
                {
                    sourceID = iii;
                    targetID = jjj;
                }


                *source_cloud_KEYpt_P_ = *lum.getPointCloud(sourceID);
                *target_cloud_KEYpt_P_ = *lum.getPointCloud(targetID);

                *source_cloud_FEATures = *FEATtVector_UNorg_P_[sourceID];
                *target_cloud_FEATures = *FEATtVector_UNorg_P_[targetID];

                // KeyPoint Estimation
                pcl::registration::CorrespondenceEstimation<keyPointTYPE,keyPointTYPE> corrEst;
                corrEst.setInputSource(            source_cloud_FEATures );
                corrEst.setInputTarget(            target_cloud_FEATures );
                corrEst.determineCorrespondences( *corrAll );

                // KeyPoint Rejection
                pcl::registration::CorrespondenceRejectorSampleConsensus<pcl::PointXYZRGBNormal> corrRej;
                corrRej.setInputSource(           source_cloud_KEYpt_P_ );
                corrRej.setInputTarget(           target_cloud_KEYpt_P_ );
                corrRej.setInlierThreshold(       PARAM_CORR_REJ_InlierThreshold     );
                corrRej.setMaximumIterations(     10000    );
                corrRej.setRefineModel(           true     );
                corrRej.setInputCorrespondences(  corrAll  );
                corrRej.getCorrespondences(      *corrFilt );

                lum.setCorrespondences( sourceID, targetID, corrFilt );

            } // if

    } // jjj

} // iii

lum.compute();

// PCLVisualizer - show this - lum.getConcatenatedCloud()

Answer 1

在对 LUM 进行了多天的试验后，我决定转向另一个用于图优化的工具，即 g2o .您可以在图像中看到结果，它并不完美(请参阅小平移漂移@正面 View )，但它合理且比简单的成对增量配准(没有非常明显的旋转漂移!)，

如果您有兴趣，我建议您下载 github version !它是最新的，而其他版本 - like this - 已经过时了，而且我个人在编译 library itself 时遇到了一些编译问题。或 my source code )