Ccny Classes

ccny_rgbd::RGBDFrame Class Reference

#include <rgbd_frame.h>

Inheritance diagram for ccny_rgbd::RGBDFrame:

clip_image001

[legend]

List of all members.

Public Member Functions

void

computeDistributions (double max_z=5.5, double max_stdev_z=0.03)

 

Computes the 3D means and covariances for all the detected keypoints.

计算所有探测关键点的均值以及方差。

void

constructFeaturePointCloud (PointCloudFeature &cloud)

 

Computes the 3D means and covariances for all the valid detected keypoints. 计算所有有效的探测关键点的均值以及方差。

EIGEN_MAKE_ALIGNED_OPERATOR_NEW

RGBDFrame ()

 

Default (empty) constructor.

 

RGBDFrame (const ImageMsg::ConstPtr &rgb_msg, const ImageMsg::ConstPtr &depth_msg, const CameraInfoMsg::ConstPtr &info_msg)

 

Constructor from ROS messages.

构造函数。

Static Public Member Functions

static bool

load (RGBDFrame &frame, const std::string &path)

 

Loads the RGBD frame from disk.

从磁盘中下载RGBD frame。

static bool

save (const RGBDFrame &frame, const std::string &path)

 

Saves the RGBD frame to disk.

保存。

Public Attributes

cv::Mat

depth_img

 

Depth image in mm (16UC1). 0 = invalid data.

16位无符号整型深度图像,0代表无效数据。

cv::Mat

descriptors

 

Feature descriptor vectors.

特征描述向量。

std_msgs::Header

header

 

Header taken from rgb_msg.

KeypointVector

keypoints

 

2D keypoint locations

Matrix3fVector

kp_covariances

 

3x3 mat of covariances

Vector3fVector

kp_means

 

1x3 mat of 3D locations

BoolVector

kp_valid

 

Is the z data valid?

image_geometry::PinholeCameraModel

model

 

The intrinsic matrix which applies to both images.

int

n_valid_keypoints

 

How many keypoints have usable 3D data.

cv::Mat

rgb_img

 

RGB image (8UC3)

Protected Member Functions

void

getGaussianDistribution (int u, int v, double &z_mean, double &z_var)

 

Calculates the z distribution (mean and variance) for a given pixel.

void

getGaussianMixtureDistribution (int u, int v, double &z_mean, double &z_var)

 

Calculates the GMM z distribution (mean and variance) for a given pixel.

double

getStdDevZ (double z)

 

Calculates the std_dev(z) from z.

double

getVarZ (double z)

 

Calculates the var(z) from z.

Static Protected Attributes

static const double

Z_STDEV_CONSTANT = 0.001425

 

Constant for calculating std_dev(z)

ccny_rgbd::RGBDKeyframe Class Reference

Extension of an RGBDFrame, which has a pose, and a 3D point cloud. More...

#include <rgbd_keyframe.h>

Inheritance diagram for ccny_rgbd::RGBDKeyframe:

clip_image002

[legend]

List of all members.

Public Member Functions

void

constructDensePointCloud (double max_z=5.5, double max_stdev_z=0.03)

 

constructs the point cloud from the RGB and depth images 从RGB和深度图像中构造点云。

EIGEN_MAKE_ALIGNED_OPERATOR_NEW

RGBDKeyframe ()

 

Default (empty) constructor.

 

RGBDKeyframe (const RGBDFrame &frame)

 

Copy constructor from a RGBDFrame.

Static Public Member Functions

static bool

load (RGBDKeyframe &keyframe, const std::string &path)

 

Loads the RGBD keyframe to disk. 从磁盘中下载RGBD

static bool

save (const RGBDKeyframe &keyframe, const std::string &path)

 

Saves the RGBD keyframe to disk.

Public Attributes

PointCloudT

cloud

 

Dense point cloud from RGBD data.

bool

manually_added

 

Whether the frame was added manually by the user. 该帧是否被用户手动添加。

double

path_length_angular

 

path length, in radians, of the camera trajectory at the moment of adding the keyframe 再添加keyframe时相机轨迹的路径长度

double

path_length_linear

 

path length, in meters, of the camera trajectory at the moment of adding the keyframe

tf::Transform

pose

 

pose of the camera, in some fixed frame

ccny_rgbd::FeatureDetector Class Reference

Inheritance diagram for ccny_rgbd::FeatureDetector:

clip_image003

Public Member Functions

 

FeatureDetector ()

 

Default constructor.

void

findFeatures (RGBDFrame &frame)

 

Main function to call to detect the sparse features in an RGBDFrame and fill out the corresponding information.

double

getMaxRange () const

 

Returns the maximum allowed z-depth (in meters) for features.

double

getMaxStDev () const

 

Returns the maximum allowed std_dev(z) (in meters) for features.

int

getSmooth () const

 

Returns the smoothing size.

void

setMaxRange (double max_range)

 

Sets the maximum allowed z-depth (in meters) for features.

void

setMaxStDev (double max_stdev)

 

Sets the maximum allowed std_dev(z) (in meters) for features.

void

setSmooth (int smooth)

 

Sets the smoothing size.

virtual

~FeatureDetector ()

 

Default destructor.

Protected Member Functions

virtual void

findFeatures (RGBDFrame &frame, const cv::Mat &input_img)=0

 

Implementation of the feature detector.

Protected Attributes

bool

compute_descriptors_

 

whether to calculate feature descriptors

boost::mutex

mutex_

 

state mutex

Private Attributes

double

max_range_

 

maximum allowed z-depth (in meters) for features

double

max_stdev_

 

maximum allowed std_dev(z) (in meters) for features

int

smooth_

 

blurring size (blur winddow = smooth*2 + 1)

Class List

Here are the classes, structs, unions and interfaces with brief descriptions:

ccny_rgbd::FeatureDetector

Base class for sparse feature extractors稀疏特征提取的基类

ccny_rgbd::FeatureHistory

Auxiliary class for the frame-to-frame ICP class 用于icp中的辅助类

ccny_rgbd::FeatureViewer

Application to test and visualize the different type of feature detectors 用于测试以及可视不同类型的特征

ccny_rgbd::GftDetector

GoodFeaturesToTrack detector

ccny_rgbd::KeyframeAssociation

Class representing an association between two keyframes, used for graph-based pose alignement 表示两关键帧之间的联系, 用于图像的配准

ccny_rgbd::KeyframeGraphDetector

Detects graph correspondences based on visual feature matching between keyframes 基于视觉特征匹配的

ccny_rgbd::KeyframeGraphSolver

Base class for graph-based global alignment classes

ccny_rgbd::KeyframeGraphSolverG2O

Graph-based global alignement using g2o (generalized graph optimizaiton)

ccny_rgbd::KeyframeMapper

Builds a 3D map from a series of RGBD keyframes 根据以系列大的关键帧构建3D 地图

ccny_rgbd::MotionEstimation

Base class for visual odometry motion estimation methods 视觉里程计运动估计方法的基类

ccny_rgbd::MotionEstimationICP

Frame-to-frame ICP motion estimation 帧到帧之间的运动估计

ccny_rgbd::MotionEstimationICPProbModel

Motion estimation based on aligning sparse features against a persistent, dynamic model 基于对应稀疏特征点的运动估计,持续,动态模型

ccny_rgbd::OrbDetector

ORB detector

ccny_rgbd::RGBDFrame

Auxiliarry class that holds together rgb and depth images

ccny_rgbd::RGBDImageProc

Processes the raw output of OpenNI sensors to create a stream of RGB-D images 处理从openni 传感器所得原始数据,得到RGB-D图像流。

ccny_rgbd::RGBDImageProcNodelet

Nodelet driver for the RGBDImageProc class

ccny_rgbd::RGBDKeyframe

Extension of an RGBDFrame, which has a pose, and a 3D point cloud RGBDFrame的扩展,包含了姿态以及3D点云。

ccny_rgbd::StarDetector

STAR detector

ccny_rgbd::SurfDetector

SURF detector

ccny_rgbd::VisualOdometry

Subscribes to incoming RGBD images and outputs the position of the moving (base) frame wrt some fixed frame 订阅RGBD图像输出运动姿态

•All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends

clip_image003[1]

clip_image001[1] clip_image004

clip_image005

clip_image006

clip_image008

clip_image010

clip_image011

clip_image012

clip_image014

转载于:https://my.oschina.net/xushizhe/blog/318946

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值