orb-slam系列 LoopClosing线程 ComputeSim3(十一)

ComputeSim3

/********************************
 * 	计算每一个回环候选关键帧与当前关键帧之间的相似矩阵
 * 		1. 首先通过BOW向量将回环候选关键帧和当前关键帧进行匹配,得到匹配地图点,通过匹配地图点初始化相似矩阵求解器
 * 		2. 遍历所有的回环候选关键帧和当前关键帧计算sim矩阵,并优化sim矩阵,根据优化sim矩阵确定匹配内点数量,从而确定此sim矩阵的准确性,以及是否可以判定为一回环.
 * 		3. 在找到的回环关键帧周围查找共视关键帧   并匹配共视关键帧的地图点和当前关键帧  相当与是匹配covisual graph和当前关键帧  根据匹配点数量确定当前帧是否发生了回环
 **************************************/
bool LoopClosing::ComputeSim3()
{
    // For each consistent loop candidate we try to compute a Sim3

    const int nInitialCandidates = mvpEnoughConsistentCandidates.size();  //候选回环帧

    // We compute first ORB matches for each candidate
    // If enough matches are found, we setup a Sim3Solver
    ORBmatcher matcher(0.75,true);
    // 相似性矩阵求解器
    vector<Sim3Solver*> vpSim3Solvers;
    vpSim3Solvers.resize(nInitialCandidates);
    // 匹配地图点容器   
    vector<vector<MapPoint*> > vvpMapPointMatches;   //vvpMapPointMatches[i][j]   候选回环帧i 对应于 待回环帧第j地图点 对应的地图点
    vvpMapPointMatches.resize(nInitialCandidates);
    // 取消候选回环关键帧的标志变量
    vector<bool> vbDiscarded;
    vbDiscarded.resize(nInitialCandidates);

    int nCandidates=0; //candidates with enough matches
    // 初始化当前帧和候选关键帧的相似矩阵求解器,计算匹配地图点,取消回环关键帧标志变量的赋初值
    for(int i=0; i<nInitialCandidates; i++)
    {
        KeyFrame* pKF = mvpEnoughConsistentCandidates[i];

        // avoid that local mapping erase it while it is being processed in this thread
        pKF->SetNotErase();

        if(pKF->isBad())
        {
            vbDiscarded[i] = true;
            continue;
        }
	// 待回环关键帧与回环候选关键帧进行匹配得到匹配地图点
        int nmatches = matcher.SearchByBoW(mpCurrentKF,pKF,vvpMapPointMatches[i]);

        if(nmatches<20)
        {
            vbDiscarded[i] = true;
            continue;
        }
        else
        {
	  //相似性矩阵的求解器初始化
            Sim3Solver* pSolver = new Sim3Solver(mpCurrentKF,pKF,vvpMapPointMatches[i],mbFixScale);
            pSolver->SetRansacParameters(0.99,20,300);
            vpSim3Solvers[i] = pSolver;
        }

        nCandidates++;
    }

    bool bMatch = false;

    // Perform alternatively RANSAC iterations for each candidate
    // until one is succesful or all fail
    // RANSAC迭代每一个回环候选关键帧和当前待回环关键帧
    // 通过SearchByBoW匹配得到初步匹配点,根据此匹配计算两关键帧之间的sim矩阵
    // 
    while(nCandidates>0 && !bMatch)
    {   
        for(int i=0; i<nInitialCandidates; i++)
        {
            if(vbDiscarded[i])  //小于20
                continue;

            KeyFrame* pKF = mvpEnoughConsistentCandidates[i];

            // Perform 5 Ransac Iterations
            vector<bool> vbInliers;
            int nInliers;
            bool bNoMore;

            Sim3Solver* pSolver = vpSim3Solvers[i];
            cv::Mat Scm  = pSolver->iterate(5,bNoMore,vbInliers,nInliers);

            // If Ransac reachs max. iterations discard keyframe   如果迭代次数达到最大  则证明当前候选帧不是回环帧
            if(bNoMore)
            {
                vbDiscarded[i]=true;
                nCandidates--;
            }

            // If RANSAC returns a Sim3, perform a guided matching and optimize with all correspondences
            // 如果RANSAC求取了一个合适的相似变换矩阵sim   则通过sim矩阵重新进行地图点匹配,并优化sim矩阵并确定内点  根据内点数量判定sim矩阵是否符合要求
            if(!Scm.empty())
            {
                vector<MapPoint*> vpMapPointMatches(vvpMapPointMatches[i].size(), static_cast<MapPoint*>(NULL));    //匹配  待回环帧i帧 对应的两地图点 =>> vpMapPointMatches[i] = loopkfi.map[j] 此帧i点 对应于 回环帧的j点
                for(size_t j=0, jend=vbInliers.size(); j<jend; j++)
                {
                    if(vbInliers[j])
                       vpMapPointMatches[j]=vvpMapPointMatches[i][j];   //第i候选回环帧 中 对应待回环帧 j 点的地图点
                }
		
                cv::Mat R = pSolver->GetEstimatedRotation();
                cv::Mat t = pSolver->GetEstimatedTranslation();
                const float s = pSolver->GetEstimatedScale();
                matcher.SearchBySim3(mpCurrentKF,pKF,vpMapPointMatches,s,R,t,7.5);

                g2o::Sim3 gScm(Converter::toMatrix3d(R),Converter::toVector3d(t),s);   //sim3  current点到 回环帧点的sim3
                const int nInliers = Optimizer::OptimizeSim3(mpCurrentKF, pKF, vpMapPointMatches, gScm, 10, mbFixScale);

                // If optimization is succesful stop ransacs and continue  
                if(nInliers>=20)
                {
                    bMatch = true;
                    mpMatchedKF = pKF;
                    g2o::Sim3 gSmw(Converter::toMatrix3d(pKF->GetRotation()),Converter::toVector3d(pKF->GetTranslation()),1.0);
                    mg2oScw = gScm*gSmw;
                    mScw = Converter::toCvMat(mg2oScw);

                    mvpCurrentMatchedPoints = vpMapPointMatches;
                    break;
                }
            }
        }
    }

    if(!bMatch)
    {
        for(int i=0; i<nInitialCandidates; i++)
             mvpEnoughConsistentCandidates[i]->SetErase();
        mpCurrentKF->SetErase();
        return false;
    }

    // Retrieve MapPoints seen in Loop Keyframe and neighbors   
    // 在找到的回环关键帧周围查找共视关键帧   并匹配共视关键帧的地图点和当前关键帧  相当于是匹配covisual graph和当前关键帧  改动关联
    vector<KeyFrame*> vpLoopConnectedKFs = mpMatchedKF->GetVectorCovisibleKeyFrames();
    vpLoopConnectedKFs.push_back(mpMatchedKF);
    mvpLoopMapPoints.clear();
    for(vector<KeyFrame*>::iterator vit=vpLoopConnectedKFs.begin(); vit!=vpLoopConnectedKFs.end(); vit++)
    {
        KeyFrame* pKF = *vit;
        vector<MapPoint*> vpMapPoints = pKF->GetMapPointMatches();     //回环帧及其共视帧的地图点
        for(size_t i=0, iend=vpMapPoints.size(); i<iend; i++)
        {
            MapPoint* pMP = vpMapPoints[i];
            if(pMP)
            {
                if(!pMP->isBad() && pMP->mnLoopPointForKF!=mpCurrentKF->mnId)
                {
                    mvpLoopMapPoints.push_back(pMP);
                    pMP->mnLoopPointForKF=mpCurrentKF->mnId;
                }
            }
        }
    }

    // Find more matches projecting with the computed Sim3  匹配共视关键帧的地图点和当前关键帧
    matcher.SearchByProjection(mpCurrentKF, mScw, mvpLoopMapPoints, mvpCurrentMatchedPoints,10);

    // If enough matches accept Loop   如果有足够的匹配点 则说明找到了回环,否则查找回环失败
    int nTotalMatches = 0;
    for(size_t i=0; i<mvpCurrentMatchedPoints.size(); i++)
    {
        if(mvpCurrentMatchedPoints[i])
            nTotalMatches++;
    }

    if(nTotalMatches>=40)
    {
        for(int i=0; i<nInitialCandidates; i++)
            if(mvpEnoughConsistentCandidates[i]!=mpMatchedKF)  //回环帧是 mpMatchedKF
                mvpEnoughConsistentCandidates[i]->SetErase();
        return true;
    }
    else
    {
        for(int i=0; i<nInitialCandidates; i++)
            mvpEnoughConsistentCandidates[i]->SetErase();
        mpCurrentKF->SetErase();
        return false;
    }

}

int nmatches = matcher.SearchByBoW(mpCurrentKF,pKF,vvpMapPointMatches[i]);

1）每一个候选回环帧都要跟待回环帧进行bow的词袋模型的匹配，并且得到匹配点。
（同一单词中两帧的关键点进行最佳匹配）
2）匹配点数大于等于20的时候进行sim3计算。

Sim3Solver* pSolver = new Sim3Solver(mpCurrentKF,pKF,vvpMapPointMatches[i],mbFixScale);

3）通过pSolver->iterate(5,bNoMore,vbInliers,nInliers); 计算出Scm
在坐标系 帧一帧二中任意选3个对点 计算sim3算法（也可以icp计算）得到位姿变化 T12
4）matcher.SearchBySim3(mpCurrentKF,pKF,vpMapPointMatches,s,R,t,7.5); 进行sim3验证

双向匹配:
根据相似矩阵将关键帧1中的地图点向关键帧2中投影,确定投影区域,并在投影区域内寻找关键帧1中地图点的匹配
根据相似矩阵将关键帧2中的地图点向关键帧1中投影,确定投影区域,并在投影区域内寻找关键帧2中地图点的匹配
根据双向匹配结果,如果两次匹配都能成功,则确定该对匹配是有效的.将其存入vpMatches12容器
最终返回匹配点对个数

帧一的地图三维点，在相机一的坐标，用优化的T12转化到相机二的坐标。验证帧二是否有这个像素点，帧二的阈值范围内找特征点匹配，找到最佳匹配。
双向 同理。
都成功是匹配点。

5）const int nInliers = Optimizer::OptimizeSim3(mpCurrentKF, pKF, vpMapPointMatches, gScm, 10, mbFixScale)
优化的策略
关键点1的uv值- K1 * t12 * p3dc2
同理 关键点2的uv值- K2 * t21 * p3dc1
优化得到sim3

6）修正

g2o::Sim3 gSmw(Converter::toMatrix3d(pKF->GetRotation()),Converter::toVector3d(pKF->GetTranslation()),1.0);
                    mg2oScw = gScm*gSmw;

因为gSmw是待回环帧在世界坐标系的位姿，计算出来当前帧 在正确世界坐标系的位姿

7）mvpCurrentMatchedPoints是当前帧和 回环帧的匹配点。mpMatchedKF是回环帧
8）mvpLoopMapPoints 回环帧及其共视帧的地图点 vpLoopConnectedKFs回环帧及其共视帧
9）matcher.SearchByProjection(mpCurrentKF, mScw, mvpLoopMapPoints, mvpCurrentMatchedPoints,10);
注意 里面3d点的来源是 mvpLoopMapPoints 也就是正确的位置3d点
p3Dc得到了在当前相机坐标系下的坐标，阈值范围内找最佳匹配 返回匹配点数目
10）数额足够 完成