<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[不能再吃了OvO的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/klsjadkls</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; klsjadkls]]></copyright><item><title><![CDATA[cv_bridgeConfig.cmake出错]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/119170761</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/119170761</guid><author>klsjadkls</author><pubDate>Wed, 28 Jul 2021 18:58:56 +0800</pubDate><description><![CDATA[https://www.cnblogs.com/xsy123/p/14488021.html
The error is about:
CMake Error at /opt/ros/melodic/share/cv_bridge/cmake/cv_bridgeConfig.cmake:113 (message):
  Project 'cv_bridge' specifies '/usr/include/opencv4/opencv' as an include
  dir, which is not fo]]></description><category></category></item><item><title><![CDATA[ORB2单目读代码笔记13--卡方检验原理及其在ORB-SLAM2中的用处]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/116735515</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/116735515</guid><author>klsjadkls</author><pubDate>Tue, 15 Jun 2021 10:05:43 +0800</pubDate><description><![CDATA[卡方检验原理及其在ORB-SLAM2中的用处FindFundamental 结束 返回 Initialize
FindFundamental 结束 返回 Initialize

    // Wait until both threads have finished
	//等待两个计算线程结束
    threadH.join();
    threadF.join();

    // Compute ratio of scores
    // Step 4 计算得分比例来判断选取哪个模型来求位姿R,t]]></description><category></category></item><item><title><![CDATA[CMAKE学习笔记]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/117429077</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/117429077</guid><author>klsjadkls</author><pubDate>Mon, 31 May 2021 20:57:00 +0800</pubDate><description><![CDATA[CMAKE官方文档：https://cmake.org/cmake/help/latest/genindex.html#
CMAKE常用指令
CMake最低版本要求
CMAKE_MINIMUM_REQUIRED( VERSION 2.8 )

项目名称
PROJECT(  项目名称  )  

定义的项目名称存放到了变量 ${PROJECT_NAME}中。
包含CMakeLists.txt所在的文件夹目录路径存放到了变量 ${PROJECT_SOURCE_DIR}中。
用   ${ } 来引用变量
设置编译]]></description><category></category></item><item><title><![CDATA[玩转 Jetson Nano——开机准备与远程连接设置]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/116781321</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/116781321</guid><author>klsjadkls</author><pubDate>Fri, 14 May 2021 15:00:40 +0800</pubDate><description><![CDATA[Nano开机准备与远程连接设置1.开机前的准备1.1 认识 Nano1.2 硬件准备1.2.1 必备1.2.2 选配1.3 在 SD 卡上烧写系统1.3.1 下载镜像1.3.2 格式化 SD 卡1.3.3 将镜像烧录到 SD 卡1.4 网络连接1.4.1 安装无线模块1.4.2 设置静态 IP1.5 供电2.远程连接2.1 远程登录终端2.1.1 获取 Nano 的 IP2.1.2 SHH 远程登录2.1.3 SSH 图形化工具 WinSCP2.2 远程登录桌面2.2.1 安装 VINO：VNC Serve]]></description><category></category></item><item><title><![CDATA[ORB2单目读代码笔记12--单目初始化中基础矩阵推导计算、根据得分确定最佳单应矩阵和基础矩阵]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115892452</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115892452</guid><author>klsjadkls</author><pubDate>Thu, 13 May 2021 09:00:22 +0800</pubDate><description><![CDATA[单目初始化中基础矩阵推导计算、根据得分确定最佳单应矩阵和基础矩阵ComputeH21 结束 返回 FindHomographyFindHomography 跳转 CheckHomography
ComputeH21 结束 返回 FindHomography
前面根据8对点（归一化）求出了单应阵H，下面反求归一化之前的
其中变量含义：

T1  T2 ：归一化矩阵。
vPn1 vPn2 ：归一化坐标。
mvKeys1 mvKeys2：原坐标。
前面求的是归一化H，即vPn1与vPn2变换的单应阵。归一化是为了]]></description><category></category></item><item><title><![CDATA[ROS多机通信SSH远程运行CCMSLAM]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/116237166</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/116237166</guid><author>klsjadkls</author><pubDate>Wed, 28 Apr 2021 17:02:30 +0800</pubDate><description><![CDATA[文章目录1.测试通信设置ROS_MASTER_URI主机rex中断输入：从机nano终端输入：启动CCMSLAM主机rex从机nano查看效果
背景：主机（笔记本，用户名rex）：安装CCMSLAM。
从机（Nano，用户名nano）：安装CCMSLAM，并可以驱动CSI摄像头，发布ROS话题。
1.测试通信
两台设备连到同一个网下，我用的是手机热点。
由于我的主机（笔记本）用的是虚拟机，需要把虚拟机网络设置为桥接模式，方法是：虚拟机 -&gt;设置 -&gt;网路适配器 -&gt;桥接模式（自动）：

在]]></description><category></category></item><item><title><![CDATA[Jetson Nano 调用CSI摄像头运行CCMSLAM]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/116207446</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/116207446</guid><author>klsjadkls</author><pubDate>Wed, 28 Apr 2021 15:09:49 +0800</pubDate><description><![CDATA[文章目录1.安装ROS的CSI摄像头软件包1.1 jetson_csi_cam1.2 jetson_nano_csi_cam2.修改Client0_euroc.launch3.编写启动文件4.启动

1.安装ROS的CSI摄像头软件包
TX2推荐安装jetson_csi_cam，而Nano推荐安装jetson_nano_csi_cam。
1.1 jetson_csi_cam
注意：与下面的jetson_nano_csi_cam二选一。
参考：https://github.com/peter-moran/je]]></description><category></category></item><item><title><![CDATA[Ubuntu 18.04 (Jetson Nano 4G/TX2)配置 CCM-SLAM]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/116116050</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/116116050</guid><author>klsjadkls</author><pubDate>Mon, 26 Apr 2021 10:35:40 +0800</pubDate><description><![CDATA[文章目录ROS
ROS
sudo apt-get update
sudo apt-get upgrade

设置最新的密钥：
sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys F42ED6FBAB17C654

安装：
sudo apt-get update
sudo apt-get install ros-melodic-desktop-full
sudo apt-get install ros-melodic-rqt*

初始化r]]></description><category></category></item><item><title><![CDATA[ORB2单目读代码笔记11--单目初始化中随机选择点集、归一化原理、直接线性变换求解单应矩阵]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115802221</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115802221</guid><author>klsjadkls</author><pubDate>Tue, 20 Apr 2021 09:28:59 +0800</pubDate><description><![CDATA[单目初始化中随机选择点集、归一化原理、直接线性变换求解单应矩阵SearchForInitialization 结束 返回 MonocularInitializationMonocularInitialization 跳转 Initialize
SearchForInitialization 结束 返回 MonocularInitialization
如果初始帧和当前帧匹配数&gt;100，则往下进行。

变量含义：
Rcw：相机旋转矩阵
tcw：相机平移向量
vbTriangulated：记录哪些特征点初始]]></description><category></category></item><item><title><![CDATA[CCM-SLAM跑自己的USB摄像头]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115856515</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115856515</guid><author>klsjadkls</author><pubDate>Mon, 19 Apr 2021 11:34:09 +0800</pubDate><description><![CDATA[CCM-SLAM跑自己的数据ccm_slam/readme.md如何使用自己的数据参数功能尝试调用usb摄像头修改摄像头启动文件`usb_cam-test.launch`测试摄像头修改Client0_euroc.launch编写启动文件启动
ccm_slam/readme.md
# 5. Using your own Data

For using you own datasets or camera, you need to create according calibration and launch ]]></description><category></category></item><item><title><![CDATA[ORB2单目读代码笔记10--单目初始化中特征匹配点对筛查]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115800590</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115800590</guid><author>klsjadkls</author><pubDate>Sat, 17 Apr 2021 21:32:50 +0800</pubDate><description><![CDATA[单目初始化中特征匹配点对筛查、隐藏bug解析修正GetFeaturesInArea 结束 返回 SearchForInitializationSearchForInitialization 结束 返回 MonocularInitialization
GetFeaturesInArea 结束 返回 SearchForInitialization
判断F2中对应位置是否有特征点，如果没有，则跳过，
如果有，则取出F1中对应特征点描述子。
        // 没有候选特征点，跳过
        if(vInd]]></description><category></category></item><item><title><![CDATA[ORB2单目读代码笔记9--单目初始化中的特征点搜索匹配、搜索候选匹配特征点]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115654124</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115654124</guid><author>klsjadkls</author><pubDate>Sat, 17 Apr 2021 20:04:58 +0800</pubDate><description><![CDATA[单目初始化中的特征点搜索匹配、搜索候选匹配特征点Frame 跳转 TrackTrack 跳转 MonocularInitializationMonocularInitialization 跳转 SearchForInitialization
Frame 跳转 Track
Track启动追踪线程
void Tracking::Track()
{
    // track包含两部分：估计运动、跟踪局部地图
    
    // mState为tracking的状态，包括 SYSTME_NOT_READY, N]]></description><category></category></item><item><title><![CDATA[[转载]卡尔曼滤波公式推导]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115799187</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115799187</guid><author>klsjadkls</author><pubDate>Sat, 17 Apr 2021 18:04:12 +0800</pubDate><description><![CDATA[转自：https://blog.youkuaiyun.com/u010720661/article/details/63253509




]]></description><category></category></item><item><title><![CDATA[[笔记]SLAM、视觉相关名词解释（遇到更新）]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115641719</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115641719</guid><author>klsjadkls</author><pubDate>Wed, 14 Apr 2021 20:15:33 +0800</pubDate><description><![CDATA[文章目录视差锚点图像梯度
视差
视差就是从有一定距离的两个点上观察同一个目标所产生的方向差异。
双目相机模型中，视差：
d=UL−UR{d = U_L - U_R}d=UL​−UR​
表示左图和右图的相机光心的距离。
锚点
指图像中，像素梯度大于阈值的点。
图像梯度
把图像看成二维离散函数，图像梯度其实就是这个二维离散函数的求导。
图像梯度: G(x,y)=dx(i,j)+dy(i,j){G(x,y) = dx(i,j) + dy(i,j)}G(x,y)=dx(i,j)+dy(i,j)
dx(i,j)=I(]]></description><category></category></item><item><title><![CDATA[ROS多机通信测试ORBSLAM2]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115707508</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115707508</guid><author>klsjadkls</author><pubDate>Wed, 14 Apr 2021 20:13:52 +0800</pubDate><description><![CDATA[文章目录ROS分布式多机通信设置ROS_MASTER_URI主机yangqing中断输入：从机rex终端输入：测试运行ORBSLAM2收尾
目标：在pc-rex下运行ORBSLAM2代码，尝试使用另外一台机器pc-yangqing显示出来。
ROS分布式多机通信
首先，使两台pc机位于同一网络中，这里我用的是手机热点。
注意：如果使用虚拟机，需要把虚拟机网络设置为桥接模式，方法是：虚拟机 -&gt;设置 -&gt;网路适配器 -&gt;桥接模式（自动）：

在两台pc上分别输入命令查看IP：
ifconfi]]></description><category></category></item><item><title><![CDATA[ubuntu安装运行ORB_SLAM2以及开启usb摄像头]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115690399</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115690399</guid><author>klsjadkls</author><pubDate>Wed, 14 Apr 2021 10:40:33 +0800</pubDate><description><![CDATA[ubuntu安装运行ORB_SLAM2以及开启usb摄像头出现问题1：error: ‘usleep’ was not declared in this scope出现问题2：ORBSLAM2中build_ros.sh编译问题解决出现问题3：ROS:使用usb_cam调用UVC摄像头遇到的坑运行结果：
转自：https://blog.youkuaiyun.com/chengmo123/article/details/104902199
出现问题1：error: ‘usleep’ was not declared in t]]></description><category></category></item><item><title><![CDATA[[转载]ROS常用命令]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115689155</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115689155</guid><author>klsjadkls</author><pubDate>Wed, 14 Apr 2021 08:16:09 +0800</pubDate><description><![CDATA[ROS常用命令创建 ROS 工作空间
转载自：https://www.cnblogs.com/liu-fa/p/5761448.html
创建 ROS 工作空间
启动 ROS
$ roscore

创建工作环境
$ mkdir -p ~/catkin_ws/src
$ cd ~/catkin_ws/src
$ catkin_init_workspace

编译 ROS 程序
$ cd ~/catkin_ws
$ catkin_make

添加程序包到全局路径
$ echo "source catkin_ws]]></description><category></category></item><item><title><![CDATA[ORB2双目读代码笔记8--两帧图像稀疏立体匹配]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115612449</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115612449</guid><author>klsjadkls</author><pubDate>Mon, 12 Apr 2021 22:33:42 +0800</pubDate><description><![CDATA[8.两帧图像稀疏立体匹配双目帧构造函数Frame(双目)Frame 跳转 ComputeStereoMatches立体匹配
双目帧构造函数
Frame(双目)
Frame::Frame(const cv::Mat &amp;imLeft, 
			const cv::Mat &amp;imRight, 
			const double &amp;timeStamp, 
			ORBextractor* extractorLeft, 
			ORBextractor* extractorRight, 
			]]></description><category></category></item><item><title><![CDATA[[笔记]keypoint.pt/.octave和Mat.ptr]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115637923</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115637923</guid><author>klsjadkls</author><pubDate>Mon, 12 Apr 2021 20:21:12 +0800</pubDate><description><![CDATA[文章目录Keypoint数据结构公有属性公有成员函数OpenCV Mat数据类型指针ptr的使用看完了这些，就能解释
Keypoint数据结构
转载自：https://www.cnblogs.com/penuel/p/13251941.html
公有属性
(float)angle：角度，表示关键点的方向，-1为初值。
(int)class_id：当要对图片进行分类时，我们可以用class_id对每个特征点进行区分，未设定时为-1，需要靠自己设定
(int)octave：代表是从金字塔哪一层提取的得到的数据。]]></description><category></category></item><item><title><![CDATA[ORB2单目读代码笔记7--去畸变、计算图像边界、特征点网格划分]]></title><link>https://blog.youkuaiyun.com/klsjadkls/article/details/115605097</link><guid>https://blog.youkuaiyun.com/klsjadkls/article/details/115605097</guid><author>klsjadkls</author><pubDate>Mon, 12 Apr 2021 19:11:39 +0800</pubDate><description><![CDATA[7.去畸变、计算图像边界、特征点网格划分7.去畸变、计算图像边界、特征点网格划分Frame 跳转 UndistortKeyPointsUndistortKeyPoints 结束 返回 Frame(疑问疑问疑问)Frame 跳转 ComputeImageBounds
7.去畸变、计算图像边界、特征点网格划分
提取图片会有广角,需要对提取到的特征点进行去畸变操作
    // Step 4 用OpenCV的矫正函数、内参对提取到的特征点进行矫正
    UndistortKeyPoints();

Frame ]]></description><category></category></item></channel></rss>