orb-slam系列 Tracking线程 orb特征提取(三)
帧对象的构造
主要功能
1）ORB extraction的入口
2）UndistortKeyPoints();
3）ComputeStereoMatches();
4）AssignFeaturesToGrid();
Frame::Frame(const cv::Mat &imLeft, const cv::Mat &imRight, const double &timeStamp, ORBextractor* extractorLeft, ORBextractor* extractorRight, ORBVocabulary* voc, cv::Mat &K, cv::Mat &distCoef, const float &bf, const float &thDepth)
:mpORBvocabulary(voc),mpORBextractorLeft(extractorLeft),mpORBextractorRight(extractorRight), mTimeStamp(timeStamp), mK(K.clone()),mDistCoef(distCoef.clone()), mbf(bf), mThDepth(thDepth),
mpReferenceKF(static_cast<KeyFrame*>(NULL))
{
// Frame ID
mnId=nNextId++;
// Scale Level Info
mnScaleLevels = mpORBextractorLeft->GetLevels();
mfScaleFactor = mpORBextractorLeft->GetScaleFactor();
mfLogScaleFactor = log(mfScaleFactor);
mvScaleFactors = mpORBextractorLeft->GetScaleFactors();
mvInvScaleFactors = mpORBextractorLeft->GetInverseScaleFactors();
mvLevelSigma2 = mpORBextractorLeft->GetScaleSigmaSquares();
mvInvLevelSigma2 = mpORBextractorLeft->GetInverseScaleSigmaSquares();
// ORB extraction
thread threadLeft(&Frame::ExtractORB,this,0,imLeft);
thread threadRight(&Frame::ExtractORB,this,1,imRight);
threadLeft.join();
threadRight.join();
N = mvKeys.size();
if(mvKeys.empty())
return;
UndistortKeyPoints();
ComputeStereoMatches();
mvpMapPoints = vector<MapPoint*>(N,static_cast<MapPoint*>(NULL));
mvbOutlier = vector<bool>(N,false);
// This is done only for the first Frame (or after a change in the calibration)
if(mbInitialComputations)
{
ComputeImageBounds(imLeft);
mfGridElementWidthInv=static_cast<float>(FRAME_GRID_COLS)/(mnMaxX-mnMinX);
mfGridElementHeightInv=static_cast<float>(FRAME_GRID_ROWS)/(mnMaxY-mnMinY);
fx = K.at<float>(0,0);
fy = K.at<float>(1,1);
cx = K.at<float>(0,2);
cy = K.at<float>(1,2);
invfx = 1.0f/fx;
invfy = 1.0f/fy;
mbInitialComputations=false;
}
mb = mbf/fx;
AssignFeaturesToGrid();
}
orb特征入口
// ORB extraction
thread threadLeft(&Frame::ExtractORB,this,0,imLeft);
thread threadRight(&Frame::ExtractORB,this,1,imRight);
/*****************************************************************
* 函数属性：类Frame的成员函数ExtractORB(int flag, const cv::Mat &im)
* 函数功能：
* 调用ORB提取器的()运算符，将得到该帧图像的关键点和描述子
* 将该帧图像的关键点存储到mvKeys
* 该帧图像的描述子存储到mDescriptors
* 函数参数介绍：
* flag：提取图像的标志 0代表左提取 1代表右提取
* im：待提取ORB特征的图像(灰度图)
* 备注：NULL
******************************************************************/
void Frame::ExtractORB(int flag, const cv::Mat &im)
{
if(flag==0)
(*mpORBextractorLeft)(im,cv::Mat(),mvKeys,mDescriptors);
else
(*mpORBextractorRight)(im,cv::Mat(),mvKeysRight,mDescriptorsRight);
}
ExtractORB函数中调用了重载的()操作符函数，该函数代码如下
void ORBextractor::operator()( InputArray _image, InputArray _mask, vector<KeyPoint>& _keypoints,
OutputArray _descriptors)
{
if(_image.empty())
return;
Mat image = _image.getMat();
assert(image.type() == CV_8UC1 );
// Pre-compute the scale pyramid 构建高斯金字塔
ComputePyramid(image);
//计算关键点并生成四叉树
vector < vector<KeyPoint> > allKeypoints;
ComputeKeyPointsOctTree(allKeypoints);
//ComputeKeyPointsOld(allKeypoints);
Mat descriptors;
int nkeypoints = 0;
for (int level = 0; level < nlevels; ++level)
nkeypoints += (int)allKeypoints[level].size();
if( nkeypoints == 0 )
_descriptors.release();
else
{
_descriptors.create(nkeypoints, 32, CV_8U);
descriptors = _descriptors.getMat();
}
_keypoints.clear();
_keypoints.reserve(nkeypoints);
//计算每个关键点对应的描述子
int offset = 0;
for (int level = 0; level < nlevels; ++level)
{
vector<KeyPoint>& keypoints = allKeypoints[level];
int nkeypointsLevel = (int)keypoints.size();
if(nkeypointsLevel==0)
continue;
// preprocess the resized image
Mat workingMat = mvImagePyramid[level].clone();
//进行高斯模糊，用BORDER_REFLECT_101方法处理边缘
GaussianBlur(workingMat, workingMat, Size(7, 7), 2, 2, BORDER_REFLECT_101);
// Compute the descriptors
Mat desc = descriptors.rowRange(offset, offset + nkeypointsLevel); //offset是偏移量，此处取出的是该层的描述子，
//计算描述子
computeDescriptors(workingMat, keypoints, desc, pattern);
offset += nkeypointsLevel;
// Scale keypoint coordinates
if (level != 0)
{
float scale = mvScaleFactor[level]; //getScale(level, firstLevel, scaleFactor);
for (vector<KeyPoint>::iterator keypoint = keypoints.begin(),
keypointEnd = keypoints.end(); keypoint != keypointEnd; ++keypoint)
keypoint->pt *= scale;
}
// And add the keypoints to the output
_keypoints.insert(_keypoints.end(), keypoints.begin(), keypoints.end());
}
}
构建高斯金金字塔
// Pre-compute the scale pyramid 构建高斯金字塔
ComputePyramid(image);
void ORBextractor::ComputePyramid(cv::Mat image)
{
for (int level = 0; level < nlevels; ++level)
{
float scale = mvInvScaleFactor[level];
Size sz(cvRound((float)image.cols*scale), cvRound((float)image.rows*scale));
// Compute the resized image
if( level != 0 )
{
resize(mvImagePyramid[level-1], mvImagePyramid[level], sz, 0, 0, INTER_LINEAR); //INTER_LINEAR是指用双线性插值的方式进行插值。
}
else
{
//此函数将图像转化为更大的图像，并用BORDER_REFLECT_101的方式来处理边缘图像
mvImagePyramid[level] = image;
}
}
}
其中mvImagePyramid[level] = image 表示该图像level的缩放图像
四叉树
vector < vector<KeyPoint> > allKeypoints;
ComputeKeyPointsOctTree(allKeypoints);
void ORBextractor::ComputeKeyPointsOctTree(vector<vector<KeyPoint> >& allKeypoints)
{
allKeypoints.resize(nlevels);
//设定每个格子的大小
const float W = 30;
//每层提想分别提取特征点
for (int level = 0; level < nlevels; ++level)
{
//设定该层图像中检测的X，Y最大最小坐标 由于PATCH 15 和EDGE_THRESHOLD 19 相差4 所以应该减去3,从每副图像的第16个元素开始查询FAST关键点
//为何要去除边缘呢，是因为边缘的特征点不具有代表性
const int minBorderX = EDGE_THRESHOLD-3;
const int minBorderY = minBorderX;
const int maxBorderX = mvImagePyramid[level].cols-EDGE_THRESHOLD+3;
const int maxBorderY = mvImagePyramid[level].rows-EDGE_THRESHOLD+3;
// 用于分配的关键点
vector<cv::KeyPoint> vToDistributeKeys;
vToDistributeKeys.reserve(nfeatures*10);
const float width = (maxBorderX-minBorderX);
const float height = (maxBorderY-minBorderY);
// 将待检测区域划分为格子的行列数
const int nCols = width/W;
const int nRows = height/W;
// 重新计算每个格子的大小
const int wCell = ceil(width/nCols);
const int hCell = ceil(height/nRows);
// 在每个格子内进行fast特征检测
for(int i=0; i<nRows; i++)
{
const float iniY =minBorderY+i*hCell;
//为什么加6呢，是因为在FAST角点检测时是需要3为半径的圆周围的16个点进行比较的，因此会加6
float maxY = iniY+hCell+6;
if(iniY>=maxBorderY-3)
continue;
if(maxY>maxBorderY)
maxY = maxBorderY;
for(int j=0; j<nCols; j++)
{
const float iniX =minBorderX+j*wCell;
float maxX = iniX+wCell+6;
if(iniX>=maxBorderX-6)
continue;
if(maxX>maxBorderX)
maxX = maxBorderX;
//每个单元格内检测
vector<cv::KeyPoint> vKeysCell;
FAST(mvImagePyramid[level].rowRange(iniY,maxY).colRange(iniX,maxX),
vKeysCell,iniThFAST,true);
//如果检测为空就降低阈值在进行检测
if(vKeysCell.empty())
{
FAST(mvImagePyramid[level].rowRange(iniY,maxY).colRange(iniX,maxX),
vKeysCell,minThFAST,true);
}
//如果检测不为空就将检测到的特征点放到vToDistributeKeys中
if(!vKeysCell.empty())
{
for(vector<cv::KeyPoint>::iterator vit=vKeysCell.begin(); vit!=vKeysCell.end();vit++)
{
(*vit).pt.x+=j*wCell;
(*vit).pt.y+=i*hCell;
vToDistributeKeys.push_back(*vit);
}
}
}
}
vector<KeyPoint> & keypoints = allKeypoints[level];
keypoints.reserve(nfeatures);
//vToDistributeKeys存储该层的关键点 minBorderX 最小x坐标 maxBorderX 最大x坐标 minBorderY最小y坐标 maxBorderY最大y坐标
//mnFeaturesPerLevel[level]该层检测到的特征点数量 level当前特征点所在的层数
keypoints = DistributeOctTree(vToDistributeKeys, minBorderX, maxBorderX,
minBorderY, maxBorderY,mnFeaturesPerLevel[level], level);
//计算特征点Patch的大小，根据每层的尺度的不同而不同
const int scaledPatchSize = PATCH_SIZE*mvScaleFactor[level];
// Add border to coordinates and scale information 将边界信息考虑进去计算特征点的位置
const int nkps = keypoints.size();
for(int i=0; i<nkps ; i++)
{
keypoints[i].pt.x+=minBorderX;//特征点的x坐标
keypoints[i].pt.y+=minBorderY;//特征点的y坐标
keypoints[i].octave=level; //特征点所在的层数
keypoints[i].size = scaledPatchSize;//特征点Patch的大小将来计算描述子时使用
}
}
// compute orientations 计算特征点的方向
for (int level = 0; level < nlevels; ++level)
computeOrientation(mvImagePyramid[level], allKeypoints[level], umax);
}
1）分为长宽30像素点的格子
2）在每个格子进行fast检测，圆周围的16个点进行比较
//每个单元格内检测
vector<cv::KeyPoint> vKeysCell;
FAST(mvImagePyramid[level].rowRange(iniY,maxY).colRange(iniX,maxX),
vKeysCell,iniThFAST,true);
//如果检测为空就降低阈值在进行检测
if(vKeysCell.empty())
{
FAST(mvImagePyramid[level].rowRange(iniY,maxY).colRange(iniX,maxX),
vKeysCell,minThFAST,true);
}
//如果检测不为空就将检测到的特征点放到vToDistributeKeys中
if(!vKeysCell.empty())
{
for(vector<cv::KeyPoint>::iterator vit=vKeysCell.begin(); vit!=vKeysCell.end();vit++)
{
(*vit).pt.x+=j*wCell;
(*vit).pt.y+=i*hCell;
vToDistributeKeys.push_back(*vit);
}
}
vit是关键点，位置赋值vToDistributeKeys中
接下来
vector<KeyPoint> & keypoints = allKeypoints[level];
keypoints.reserve(nfeatures);
//vToDistributeKeys存储该层的关键点 minBorderX 最小x坐标 maxBorderX 最大x坐标 minBorderY最小y坐标 maxBorderY最大y坐标
//mnFeaturesPerLevel[level]该层检测到的特征点数量 level当前特征点所在的层数
keypoints = DistributeOctTree(vToDistributeKeys, minBorderX, maxBorderX,
minBorderY, maxBorderY,mnFeaturesPerLevel[level], level);
DistributeOctTree()函数 四分树
vector<cv::KeyPoint> ORBextractor::DistributeOctTree(const vector<cv::KeyPoint>& vToDistributeKeys, const int &minX,
const int &maxX, const int &minY, const int &maxY, const int &N, const int &level)
{
// Compute how many initial nodes 计算有多少个初始的节点数量，为了满足每个节点是正方形的，根节点的数量
const int nIni = round(static_cast<float>(maxX-minX)/(maxY-minY));
//得到每个节点之间的距离
const float hX = static_cast<float>(maxX-minX)/nIni;
//分开的块
list<ExtractorNode> lNodes;
vector<ExtractorNode*> vpIniNodes;
vpIniNodes.resize(nIni);
//初始化各个节点
for(int i=0; i<nIni; i++)
{
//四叉树是每次根据特定条件将一个结点分成四等分，四个区域左上(UL)，右上(UR)，左下(BL)，右下(BR
ExtractorNode ni;
ni.UL = cv::Point2i(hX*static_cast<float>(i),0);
ni.UR = cv::Point2i(hX*static_cast<float>(i+1),0);
ni.BL = cv::Point2i(ni.UL.x,maxY-minY);
ni.BR = cv::Point2i(ni.UR.x,maxY-minY);
ni.vKeys.reserve(vToDistributeKeys.size());
lNodes.push_back(ni);
vpIniNodes[i] = &lNodes.back();
}
//Associate points to childs 把关键点分配到各个节点中
for(size_t i=0;i<vToDistributeKeys.size();i++)
{
const cv::KeyPoint &kp = vToDistributeKeys[i];
vpIniNodes[kp.pt.x/hX]->vKeys.push_back(kp);
}
list<ExtractorNode>::iterator lit = lNodes.begin();
while(lit!=lNodes.end())
{
//如果节点中只有一个特征点时，不在进行划分
if(lit->vKeys.size()==1)
{
lit->bNoMore=true;
lit++;
}
else if(lit->vKeys.empty()) //如果该节点中无特征点时，就删除该节点
lit = lNodes.erase(lit);
else
lit++;
}
bool bFinish = false;
int iteration = 0;
vector<pair<int,ExtractorNode*> > vSizeAndPointerToNode;//节点及对应包含的特征数
vSizeAndPointerToNode.reserve(lNodes.size()*4);
while(!bFinish)
{
iteration++;
// 初始节点个数，用于判断是否节点再一次进行了划分
int prevSize = lNodes.size();
lit = lNodes.begin();
// 表示节点分解次数
int nToExpand = 0;
vSizeAndPointerToNode.clear();
while(lit!=lNodes.end())
{
if(lit->bNoMore)
{
// If node only contains one point do not subdivide and continue 表面只有一个特征点，不在进行划分
lit++;
continue;
}
else
{
// If more than one point, subdivide 如果有超过一个节点就继续划分
ExtractorNode n1,n2,n3,n4;
lit->DivideNode(n1,n2,n3,n4);//将该节点划分为4个节点
// Add childs if they contain points 如果子节点中包含着特征点就添加该节点，否则删除节点
if(n1.vKeys.size()>0)
{
lNodes.push_front(n1);
if(n1.vKeys.size()>1)
{
nToExpand++;
vSizeAndPointerToNode.push_back(make_pair(n1.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
if(n2.vKeys.size()>0)
{
lNodes.push_front(n2);
if(n2.vKeys.size()>1)
{
nToExpand++;
vSizeAndPointerToNode.push_back(make_pair(n2.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
if(n3.vKeys.size()>0)
{
lNodes.push_front(n3);
if(n3.vKeys.size()>1)
{
nToExpand++;
vSizeAndPointerToNode.push_back(make_pair(n3.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
if(n4.vKeys.size()>0)
{
lNodes.push_front(n4);
if(n4.vKeys.size()>1)
{
nToExpand++;
vSizeAndPointerToNode.push_back(make_pair(n4.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
//删除该节点，因为有字节点包含了这些特征点了，所以为了减少冗余节点就删除该父节点
lit=lNodes.erase(lit);
continue;
}
}
// Finish if there are more nodes than required features
// or all nodes contain just one point// 当节点个数大于需分配的特征数或者所有的节点只有一个特征点（节点不能划分）的时候，则结束。
if((int)lNodes.size()>=N || (int)lNodes.size()==prevSize)
{
bFinish = true;
}
else if(((int)lNodes.size()+nToExpand*3)>N)//节点展开次数乘以3用于表明下一次的节点分解可能超过特征数，即为最后一次分解
{
while(!bFinish)
{
prevSize = lNodes.size();
vector<pair<int,ExtractorNode*> > vPrevSizeAndPointerToNode = vSizeAndPointerToNode;
vSizeAndPointerToNode.clear();
sort(vPrevSizeAndPointerToNode.begin(),vPrevSizeAndPointerToNode.end());
for(int j=vPrevSizeAndPointerToNode.size()-1;j>=0;j--)
{
ExtractorNode n1,n2,n3,n4;
vPrevSizeAndPointerToNode[j].second->DivideNode(n1,n2,n3,n4);
// Add childs if they contain points // 划分之后进一步的判断
if(n1.vKeys.size()>0)
{
lNodes.push_front(n1);
if(n1.vKeys.size()>1)
{
vSizeAndPointerToNode.push_back(make_pair(n1.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
if(n2.vKeys.size()>0)
{
lNodes.push_front(n2);
if(n2.vKeys.size()>1)
{
vSizeAndPointerToNode.push_back(make_pair(n2.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
if(n3.vKeys.size()>0)
{
lNodes.push_front(n3);
if(n3.vKeys.size()>1)
{
vSizeAndPointerToNode.push_back(make_pair(n3.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
if(n4.vKeys.size()>0)
{
lNodes.push_front(n4);
if(n4.vKeys.size()>1)
{
vSizeAndPointerToNode.push_back(make_pair(n4.vKeys.size(),&lNodes.front()));
lNodes.front().lit = lNodes.begin();
}
}
lNodes.erase(vPrevSizeAndPointerToNode[j].second->lit);
if((int)lNodes.size()>=N)
break;
}
if((int)lNodes.size()>=N || (int)lNodes.size()==prevSize)
bFinish = true;
}
}
}
// Retain the best point in each node 保留每个节点下最好的特征点
vector<cv::KeyPoint> vResultKeys;
vResultKeys.reserve(nfeatures);
for(list<ExtractorNode>::iterator lit=lNodes.begin(); lit!=lNodes.end(); lit++)
{
vector<cv::KeyPoint> &vNodeKeys = lit->vKeys;
cv::KeyPoint* pKP = &vNodeKeys[0];
float maxResponse = pKP->response;
for(size_t k=1;k<vNodeKeys.size();k++)
{
if(vNodeKeys[k].response>maxResponse)
{
pKP = &vNodeKeys[k];
maxResponse = vNodeKeys[k].response;
}
}
vResultKeys.push_back(*pKP);
}
return vResultKeys;
}
const int nIni = round(static_cast<float>(maxX-minX)/(maxY-minY));
四舍五入根节点 正方形
for(int i=0; i<nIni; i++)
{
//四叉树是每次根据特定条件将一个结点分成四等分，四个区域左上(UL)，右上(UR)，左下(BL)，右下(BR
ExtractorNode ni;
ni.UL = cv::Point2i(hX*static_cast<float>(i),0);
ni.UR = cv::Point2i(hX*static_cast<float>(i+1),0);
ni.BL = cv::Point2i(ni.UL.x,maxY-minY);
ni.BR = cv::Point2i(ni.UR.x,maxY-minY);
ni.vKeys.reserve(vToDistributeKeys.size());
lNodes.push_back(ni);
vpIniNodes[i] = &lNodes.back();
}
//Associate points to childs 把关键点分配到各个节点中
for(size_t i=0;i<vToDistributeKeys.size();i++)
{
const cv::KeyPoint &kp = vToDistributeKeys[i];
vpIniNodes[kp.pt.x/hX]->vKeys.push_back(kp);
}
不断分成四分 知道每个小框中只有一个节点
vResultKeys.push_back(*pKP) 存入返回
computeOrientation 计算方向
// compute orientations 计算特征点的方向
for (int level = 0; level < nlevels; ++level)
computeOrientation(mvImagePyramid[level], allKeypoints[level], umax);
static void computeOrientation(const Mat& image, vector<KeyPoint>& keypoints, const vector<int>& umax)
{
for (vector<KeyPoint>::iterator keypoint = keypoints.begin(),
keypointEnd = keypoints.end(); keypoint != keypointEnd; ++keypoint)
{
keypoint->angle = IC_Angle(image, keypoint->pt, umax);
}
}
// 灰度质心法计算特征点方向 image待提取特征点的图片 pt是特征点的位置
static float IC_Angle(const Mat& image, Point2f pt, const vector<int> & u_max)
{
int m_01 = 0, m_10 = 0;
const uchar* center = &image.at<uchar> (cvRound(pt.y), cvRound(pt.x));
// Treat the center line differently, v=0 v=0这一行是特殊的，需要单独计算
for (int u = -HALF_PATCH_SIZE; u <= HALF_PATCH_SIZE; ++u)
m_10 += u * center[u];//m_10=SUM(u*I(u,v)) m_01=SUM(v*I(u,v))
// Go line by line in the circuI853lar patch
int step = (int)image.step1();
for (int v = 1; v <= HALF_PATCH_SIZE; ++v)
{
// Proceed over the two lines
int v_sum = 0;
int d = u_max[v];
for (int u = -d; u <= d; ++u)
{
int val_plus = center[u + v*step], val_minus = center[u - v*step];
v_sum += (val_plus - val_minus); //计算上下的时候是有符号的，所以这边是减
m_10 += u * (val_plus + val_minus);//将(u,v)和(u,-v)两个点的像素一起计算 这边加是由于u已经确定好了符号
}
m_01 += v * v_sum; //将x=v这条直线上所有的坐标点的像素值求和在进行计算
}
return fastAtan2((float)m_01, (float)m_10);
}
灰度质心法
computeOrbDescriptor
static void computeOrbDescriptor(const KeyPoint& kpt,
const Mat& img, const Point* pattern,
uchar* desc)
{
//取关键点的方向角度并进一步转化为弧度
float angle = (float)kpt.angle*factorPI;
float a = (float)cos(angle), b = (float)sin(angle);
const uchar* center = &img.at<uchar>(cvRound(kpt.pt.y), cvRound(kpt.pt.x));
const int step = (int)img.step;
//使得该关键点具有旋转不变性
#define GET_VALUE(idx) \
center[cvRound(pattern[idx].x*b + pattern[idx].y*a)*step + \
cvRound(pattern[idx].x*a - pattern[idx].y*b)]
//生成8*32 = 256维的向量代表这个关键点，并通过将点的坐标转换到以描述子主方向为X轴的坐标系下，从而使得描述子具有旋转不变性。
for (int i = 0; i < 32; ++i, pattern += 16)
{
int t0, t1, val;
t0 = GET_VALUE(0); t1 = GET_VALUE(1);
val = t0 < t1;
t0 = GET_VALUE(2); t1 = GET_VALUE(3);
val |= (t0 < t1) << 1;
t0 = GET_VALUE(4); t1 = GET_VALUE(5);
val |= (t0 < t1) << 2;
t0 = GET_VALUE(6); t1 = GET_VALUE(7);
val |= (t0 < t1) << 3;
t0 = GET_VALUE(8); t1 = GET_VALUE(9);
val |= (t0 < t1) << 4;
t0 = GET_VALUE(10); t1 = GET_VALUE(11);
val |= (t0 < t1) << 5;
t0 = GET_VALUE(12); t1 = GET_VALUE(13);
val |= (t0 < t1) << 6;
t0 = GET_VALUE(14); t1 = GET_VALUE(15);
val |= (t0 < t1) << 7;
desc[i] = (uchar)val;
}
#undef GET_VALUE
}
注意角度