📄 文章列表

备份文件
cp ~/slam_ws/ORB_SLAM3/Examples/Monocular/mono_tum.cc \
   ~/slam_ws/ORB_SLAM3/Examples/Monocular/mono_tum_backup.cc
还原文件
cp mono_tum_backup.cc mono_tum.cc

备份文件
cp -r ~/slam_ws/ORB_SLAM3 ~/slam_ws/ORB_SLAM3_backup

cd ~/slam_ws/ORB_SLAM3
python3 -m venv evo_env
source evo_env/bin/activate
pip install --upgrade pip
pip install "numpy<1.25" "scipy>=1.8,<1.11" evo

安装QT6,防止报错
pip install PyQt6

就补系统库
apt update
apt install -y libxcb-cursor0 libxkbcommon-x11-0 libgl1

配置yolo环境
cd ~/slam_ws
python3 -m venv yolo_env
source yolo_env/bin/activate
pip install --upgrade pip
pip install torch==2.2.2 torchvision==0.17.2 --index-url https://download.pytorch.org/whl/cu121
pip install ultralytics opencv-python

工作空间初始化
cd ~/colcon_ws
mkdir -p src
colcon build
source install/setup.bash

准备一个 SLAM 专用目录
mkdir -p ~/slam_ws
cd ~/slam_ws
安装依赖
sudo apt update
sudo apt install -y \
  git cmake build-essential pkg-config \
  libeigen3-dev \
  libopencv-dev \
  libglew-dev \
  libboost-dev libboost-serialization-dev libboost-system-dev libboost-filesystem-dev \
  libpython3-dev python3-numpy \
  ffmpeg libavcodec-dev libavutil-dev libavformat-dev libswscale-dev libavdevice-dev \
  libjpeg-dev libpng-dev libtiff5-dev libopenexr-dev

sudo apt install -y libepoxy-dev

✅ 稳定版 Pangolin
cd ~/slam_ws
git clone https://github.com/stevenlovegrove/Pangolin.git
cd Pangolin

git checkout v0.6

✅ 第一步：用 nano 打开文件
nano ~/slam_ws/Pangolin/include/pangolin/gl/colour.h
✅ 第二步：找到 include 区域
你会看到文件开头类似：
#include <pangolin/gl/gl.h>
✅ 第三步：加这一行
在这些 #include 下面加一行：
#include <limits>
✅ 第四步：编译
mkdir build && cd build
cd ~/slam_ws/Pangolin/build
make clean
cmake ..
make -j"$(nproc)"
sudo make install
sudo ldconfig

cd ~/slam_ws
git clone https://github.com/UZ-SLAMLab/ORB_SLAM3.git
cd ORB_SLAM3
chmod +x build.sh
./build.sh

mkdir -p ~/dataset
cd ~/dataset
wget https://cvg.cit.tum.de/rgbd/dataset/freiburg1/rgbd_dataset_freiburg1_xyz.tgz
tar -xvzf rgbd_dataset_freiburg1_xyz.tgz

set +e

# 只要这个终端窗口准备关闭（无论是因为报错还是输入了 exit），强制停下读一个回车
trap 'echo -e "\n\033[31m[检测到终端尝试退出]\033[0m"; read -p "进程已结束，按回车键才准关闭窗口..." ' EXIT

docker run -it -d \
  --name code-server-cuda \
  --runtime nvidia \
  --gpus all \
  --shm-size=16gb \
  --security-opt seccomp=unconfined \
  --cap-add=SYS_PTRACE \
  -e NVIDIA_VISIBLE_DEVICES=all \
  -e NVIDIA_DRIVER_CAPABILITIES=compute,graphics,display,utility,video \
  -e DISPLAY=:99 \
  -p 8080:8080 \
  -p 1000:6080 \
  -p 1001:5900 \
  -p 1002:8888 \
  -p 1003:80 \
  -p 1004:443 \
  -p 1005:3000 \
  -v "/home/nvidia/taospace/cuda-code-server/config:/home/coder/.config" \
  -v "/home/nvidia/taospace/cuda-code-server/project:/home/coder/project" \
  --restart no \
  dustynv/ros:humble-desktop-pytorch-l4t-r35.4.1 \
  /bin/bash

nvidia@nvidia-desktop:~$ head -n 1 /etc/nv_tegra_release
# R35 (release), REVISION: 4.1, GCID: 33958178, BOARD: t186ref, EABI: aarch64, DATE: Tue Aug  1 19:57:35 UTC 2023
nvidia@nvidia-desktop:~$ sudo apt-cache show nvidia-jetpack | grep Version
[sudo] nvidia 的密码： 
Version: 5.1.2-b104
nvidia@nvidia-desktop:~$

# 创建虚拟环境 (名为 .venv)（下面保留主环境库）
python3 -m venv .venv
python3 -m venv --system-site-packages .venv
# 激活环境
source .venv/bin/activate
# 退出激活
deactivate

services:
  ros2_slam:
    image: ros:humble-perception
    container_name: ros2_slam
    tty: true
    stdin_open: true
    network_mode: host
    ipc: host
    privileged: true
    volumes:
      - /dev:/dev
      - ./ros2_ws:/root/ros2_ws
    environment:
      - ROS_DOMAIN_ID=0
      - RMW_IMPLEMENTATION=rmw_fastrtps_cpp
    working_dir: /root
    command: bash

docker run -d \
  --name code-server-gpu \
  --runtime nvidia \
  --gpus all \
  --security-opt seccomp=unconfined \
  -p 8080:8080 \
  -p 1000:6080 \
  -p 1001:5900 \
  -p 1002:8888 \
  -p 1003:80 \
  -p 1004:443 \
  -p 1005:3000 \
  -v "/home/nvidia/taospace/code-server/config:/home/coder/.config" \
  -v "/home/nvidia/taospace/code-server/project:/home/coder/project" \
  -e AUTH=password \
  -e PASSWORD=password \
  -e DISPLAY=:99 \
  -u "$(id -u):$(id -g)" \
  --restart always \
  codercom/code-server:latest \
  --cert

docker run -it -d \
  --name ros_gpu_final \
  --runtime nvidia \
  --gpus all \
  --network host \
  --shm-size=16gb \
  -e NVIDIA_VISIBLE_DEVICES=all \
  -e NVIDIA_DRIVER_CAPABILITIES=compute,graphics,display,utility,video \
  -e DISPLAY=:99 \
  dustynv/ros:humble-desktop-pytorch-l4t-r35.4.1