<blockquote data-ke-style="style3"><b>1. Stereo Camera > 3D map (Visual SLAM)</b><br><b>2. AI로 객체 인지</b><br><b>3. Jetson 환경에 임베딩</b><br> <br>이 세 가지를 합쳐서 작품을 구성해보려고 함<br>3D LiDAR 사용 없이 3D 재구성 + Segmentation<br><br>https://cafe.daum.net/SmartRobot/RoVa/2424</blockquote><hr data-ke-style="style6"><p><span data-ke-size="size18"><b>1. RTAB-Map을 carla 시뮬레이터에서 응용한 사례가 있는지 조사</b></span></p><p> </p><p><a href="https://youtu.be/jTs--Fx2sxQ?si=GYgloWdkta7pHFWg" target="_blank" class="ke-link">https://youtu.be/jTs--Fx2sxQ?si=GYgloWdkta7pHFWg</a></p><div class="figure-video" data-ke-type="video" data-video-url="https://www.youtube.com/watch?v=jTs--Fx2sxQ" data-video-host="youtube" data-video-play-service="daum_cafe" data-video-width="868" data-video-height="488" data-ke-style="alignLeft" data-video-thumbnail="https://scrap.kakaocdn.net/dn/2ObpI/hyZOKWAc4Q/k8cte7jzyiPLx3E9QYQvjK/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/dQlFzL/hyZOGNq48K/udEvbBOAje1uNizK3iOodk/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/5mlX8/hyZO1KWNgG/GGaCtSyp6N5F4Kdv1iqm9K/img.jpg?width=1280&height=720&face=0_0_1280_720" data-ke-mobile-mobilestyle="widthContent" data-video-origin-width="868" data-video-origin-height="488"><iframe src="https://www.youtube.com/embed/jTs--Fx2sxQ" width="868" height="488" frameborder="0" scrolling="no" allowfullscreen=""></iframe><div class="figcaption"></div></div><p> </p><p><a href="https://youtu.be/6ewC3qSw9zM?si=tXu2Ev54lifm6BuJ" target="_blank" class="ke-link">https://youtu.be/6ewC3qSw9zM?si=tXu2Ev54lifm6BuJ</a></p><div class="figure-video" data-ke-type="video" data-video-url="https://www.youtube.com/watch?v=6ewC3qSw9zM" data-video-host="youtube" data-video-play-service="daum_cafe" data-video-width="853" data-video-height="480" data-ke-style="alignCenter" data-video-thumbnail="https://scrap.kakaocdn.net/dn/bPlwzD/hyZOZM8ykB/HVeJwAEHU777Eq4WZJBOK1/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/cj6WWm/hyZO4OrZ8y/T0PHZM4DpDue6rETPFMgz1/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/cOLNZW/hyZOWCTmSG/gWbvmlvhpaOT4FpkzAA4R1/img.jpg?width=1280&height=720&face=0_0_1280_720" data-ke-mobile-mobilestyle="widthContent" data-video-origin-width="853" data-video-origin-height="480"><iframe src="https://www.youtube.com/embed/6ewC3qSw9zM" width="853" height="480" frameborder="0" scrolling="no" allowfullscreen=""></iframe><div class="figcaption"></div></div><p> </p><p>CARLA는 센서 데이터를 ROS 2 토픽으로 쏴주는 <b>carla-ros-bridge</b> 라는 패키지를 제공</p><p> </p><p>ROS 2 브릿지를 켬 >>> RTAB-Map 노드를 켜서 해당 토픽을 구독(Subscribe)하게 설정</p><p>차량이 도시를 주행하면 RTAB-Map이 실시간으로 3D 지도를 만듦</p><p> </p><p> </p><p> </p><hr data-ke-style="style6"><p> </p><p> </p><p><span data-ke-size="size18"><b>2. RTAB-Map을 gazebo 시뮬레이터에서 응용한 사례가 있는지 조사</b></span></p><p> </p><p><a href="https://youtu.be/2rAXuFpskYc?si=GiI3MZy5L4ZPxz7M" target="_blank" class="ke-link">https://youtu.be/2rAXuFpskYc?si=GiI3MZy5L4ZPxz7M</a></p><div class="figure-video" data-ke-type="video" data-video-url="https://www.youtube.com/watch?v=2rAXuFpskYc" data-video-host="youtube" data-video-play-service="daum_cafe" data-video-width="853" data-video-height="480" data-ke-style="alignCenter" data-video-thumbnail="https://scrap.kakaocdn.net/dn/mc8Zv/hyZO1jUzn9/IIkkzTMvVe4fve8FzcdcS1/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/c5HXLW/hyZOWXaITS/BWYSMQpmMPUKAmDWlUC8KK/img.jpg?width=1280&height=720&face=0_0_1280_720" data-ke-mobile-mobilestyle="widthContent" data-video-origin-width="853" data-video-origin-height="480"><iframe src="https://www.youtube.com/embed/2rAXuFpskYc" width="853" height="480" frameborder="0" scrolling="no" allowfullscreen=""></iframe><div class="figcaption"></div></div><p> </p><p><a href="https://youtu.be/9GBzygVMX-k?si=Lb6SDkFeP2TCfkCt" target="_blank" class="ke-link">https://youtu.be/9GBzygVMX-k?si=Lb6SDkFeP2TCfkCt</a></p><div class="figure-video" data-ke-type="video" data-video-url="https://www.youtube.com/watch?v=9GBzygVMX-k" data-video-host="youtube" data-video-play-service="daum_cafe" data-video-width="853" data-video-height="480" data-ke-style="alignCenter" data-video-thumbnail="https://scrap.kakaocdn.net/dn/f8gEK/hyZO4AVV4o/GTSHAiKrtknkU6TXwkEP0k/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/cW7vA7/hyZNxduIFn/9qbyOdw0YrTf6siTjS4YK0/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/bfMbYe/hyZOWizOxx/SGoIxm7ddK9mL7wAgnLKu0/img.jpg?width=1280&height=720&face=0_0_1280_720" data-ke-mobile-mobilestyle="widthContent" data-video-origin-width="853" data-video-origin-height="480"><iframe src="https://www.youtube.com/embed/9GBzygVMX-k" width="853" height="480" frameborder="0" scrolling="no" allowfullscreen=""></iframe><div class="figcaption"></div></div><p> </p><p>거의 모든<b> RTAB-Map 튜토리얼이 Gazebo를 기반</b>으로 함</p><p> </p><p>TurtleBot3 시뮬레이션을 켜고 RTAB-Map을 실행하는 예제</p><p><a href="https://wiki.ros.org/rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot" target="_blank" class="ke-link">https://wiki.ros.org/rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot</a></p><div class="figure-open" contenteditable="false" data-ke-type="opengraph" data-ke-align="alignCenter" data-og-type="website" data-og-title="rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot - ROS Wiki" data-og-description="Please ask about problems and questions regarding this tutorial on answers.ros.org. Don't forget to include in your question the link to this page, the versions of your OS & ROS, and also add appropriate tags. Mapping and Navigation with Turtlebot Descript" data-og-host="wiki.ros.org" data-og-source-url="https://wiki.ros.org/rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot" data-og-url="https://wiki.ros.org/rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot" data-og-image="https://scrap.kakaocdn.net/dn/cLBpB8/hyZNBGXGeu/wNKX0ztRuUQZYac4mIKLhk/img.png?width=750&height=550&face=0_0_750_550,https://scrap.kakaocdn.net/dn/bAxG75/hyZOXPkLBt/R0ZM2iAM2RjrdbREaRfnpK/img.png?width=750&height=410&face=0_0_750_410"><a href="https://wiki.ros.org/rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot" target="_blank" data-source-url="https://wiki.ros.org/rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot"><div class="og-image"><img class="thumb_img" src="https://scrap.kakaocdn.net/dn/cLBpB8/hyZNBGXGeu/wNKX0ztRuUQZYac4mIKLhk/img.png?width=750&height=550&face=0_0_750_550,https://scrap.kakaocdn.net/dn/bAxG75/hyZOXPkLBt/R0ZM2iAM2RjrdbREaRfnpK/img.png?width=750&height=410&face=0_0_750_410" alt="" xxxxonerror="this.src="//img1.kakaocdn.net/thumb/C200x200/?fname=https%3A%2F%2Ft1.daumcdn.net%2Fcafe_image%2Fcafe_meta_image_190529.png""></div><div class="og-text"><p class="og-title">rtabmap_ros/Tutorials/MappingAndNavigationOnTurtlebot - ROS Wiki</p><p class="og-desc">Please ask about problems and questions regarding this tutorial on answers.ros.org. Don't forget to include in your question the link to this page, the versions of your OS & ROS, and also add appropriate tags. Mapping and Navigation with Turtlebot Descript</p><p class="og-host">wiki.ros.org</p></div></a></div><p> </p><p> </p><hr data-ke-style="style6"><p> </p><p><span data-ke-size="size18"><b>3. RTAB-Map의 localization 기능 조사 (slam 후에 맵을 이용하여 위치 추정하는건지?)</b></span><b></b></p><p> </p><p>slam 후에 맵을 이용하여 위치 추정을 함 == <b>Localization Mode</b></p><p> </p><ul style="list-style-type: disc;" data-ke-list-type="disc"><li><b>Mapping Mode (지도 작성 모드..)</b><br><br>동작: <b>지도를 계속 추가하고 확장</b>함, 메모리 늘어남</li><li><b>Localization Mode (위치 추정 모드)</b><br><br>동작: 이미 다 그려진 지도를 불러옴, 새로운 데이터를 지도에 추가하지 않음<br>원리: 카메라로 현재 장면을 봄 > <b>저장된 지도 데이터베이스에서 장면과 같은 곳은 어딘지 검색</b>함<br>(Visual Loop Closure), 매칭되면 위치를 잡음</li></ul><p> </p><p> </p><hr data-ke-style="style6"><p> </p><p> </p><p><span data-ke-size="size18"><b>4. slam을 수행하는 AI 모델이 있는지 조사</b></span></p><p> </p><p>slam을 수행하는 AI 모델을 <b>Deep Learning Based SLAM</b>이라고 함</p><p> </p><p><b>End-to-End 방식의 SLAM: Droid-SLAM, TartanVO</b></p><p><span style="color: #333333; background-color: #ffffff;" data-ke-size="size16">>>> </span><b>이미지를 넣으면 뉴럴 네트워크가 알아서 이동했음을 인지하고 포즈을 출력으로 줌</b></p><p><a href="https://github.com/princeton-vl/DROID-SLAM" target="_blank" class="ke-link">https://github.com/princeton-vl/DROID-SLAM</a></p><div class="figure-open" contenteditable="false" data-ke-type="opengraph" data-ke-align="alignCenter" data-og-type="website" data-og-title="GitHub - princeton-vl/DROID-SLAM" data-og-description="Contribute to princeton-vl/DROID-SLAM development by creating an account on GitHub." data-og-host="github.com" data-og-source-url="https://github.com/princeton-vl/DROID-SLAM" data-og-url="https://github.com/princeton-vl/DROID-SLAM" data-og-image="https://scrap.kakaocdn.net/dn/D1Ve7/hyZNLbLyRS/RFAlyIeatevTXdWikKrX41/img.png?width=1200&height=600&face=0_0_1200_600,https://scrap.kakaocdn.net/dn/b4Yqis/hyZNEDETgs/RgkRB8g4GhZAy0Sie85of1/img.png?width=1200&height=600&face=0_0_1200_600"><a href="https://github.com/princeton-vl/DROID-SLAM" target="_blank" data-source-url="https://github.com/princeton-vl/DROID-SLAM"><div class="og-image"><img class="thumb_img" src="https://scrap.kakaocdn.net/dn/D1Ve7/hyZNLbLyRS/RFAlyIeatevTXdWikKrX41/img.png?width=1200&height=600&face=0_0_1200_600,https://scrap.kakaocdn.net/dn/b4Yqis/hyZNEDETgs/RgkRB8g4GhZAy0Sie85of1/img.png?width=1200&height=600&face=0_0_1200_600" alt="" xxxxonerror="this.src="//img1.kakaocdn.net/thumb/C200x200/?fname=https%3A%2F%2Ft1.daumcdn.net%2Fcafe_image%2Fcafe_meta_image_190529.png""></div><div class="og-text"><p class="og-title">GitHub - princeton-vl/DROID-SLAM</p><p class="og-desc">Contribute to princeton-vl/DROID-SLAM development by creating an account on GitHub.</p><p class="og-host">github.com</p></div></a></div><p><a href="https://youtu.be/GG78CSlSHSA?si=GnAIcLKATWCnDYvr" target="_blank" class="ke-link">https://youtu.be/GG78CSlSHSA?si=GnAIcLKATWCnDYvr</a></p><div class="figure-video" data-ke-type="video" data-video-url="https://www.youtube.com/watch?v=GG78CSlSHSA" data-video-host="youtube" data-video-play-service="daum_cafe" data-video-width="853" data-video-height="480" data-ke-style="alignCenter" data-video-thumbnail="https://scrap.kakaocdn.net/dn/bNVYze/hyZOTsByyk/8omDPW88NP5au2iOVOBqG0/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/cdPvfL/hyZOTlRV3P/RHHeyREkJEudjpWmBI09ik/img.jpg?width=1280&height=720&face=0_0_1280_720" data-ke-mobile-mobilestyle="widthContent" data-video-origin-width="853" data-video-origin-height="480"><iframe src="https://www.youtube.com/embed/GG78CSlSHSA" width="853" height="480" frameborder="0" scrolling="no" allowfullscreen=""></iframe><div class="figcaption"></div></div><p> </p><p> </p><p><a href="https://github.com/castacks/tartanvo" target="_blank" class="ke-link">https://github.com/castacks/tartanvo</a></p><div class="figure-open" contenteditable="false" data-ke-type="opengraph" data-ke-align="alignCenter" data-og-type="website" data-og-title="GitHub - castacks/tartanvo: TartanVO: A Generalizable Learning-based VO" data-og-description="TartanVO: A Generalizable Learning-based VO. Contribute to castacks/tartanvo development by creating an account on GitHub." data-og-host="github.com" data-og-source-url="https://github.com/castacks/tartanvo" data-og-url="https://github.com/castacks/tartanvo" data-og-image="https://scrap.kakaocdn.net/dn/vJv7y/hyZO48MXSm/XrBVbaiukTTtexJXkiZAfK/img.png?width=1200&height=600&face=0_0_1200_600,https://scrap.kakaocdn.net/dn/J0aIe/hyZO3WlwvH/hssGlFsooEIaQ3iRbb2e20/img.png?width=1200&height=600&face=0_0_1200_600"><a href="https://github.com/castacks/tartanvo" target="_blank" data-source-url="https://github.com/castacks/tartanvo"><div class="og-image"><img class="thumb_img" src="https://scrap.kakaocdn.net/dn/vJv7y/hyZO48MXSm/XrBVbaiukTTtexJXkiZAfK/img.png?width=1200&height=600&face=0_0_1200_600,https://scrap.kakaocdn.net/dn/J0aIe/hyZO3WlwvH/hssGlFsooEIaQ3iRbb2e20/img.png?width=1200&height=600&face=0_0_1200_600" alt="" xxxxonerror="this.src="//img1.kakaocdn.net/thumb/C200x200/?fname=https%3A%2F%2Ft1.daumcdn.net%2Fcafe_image%2Fcafe_meta_image_190529.png""></div><div class="og-text"><p class="og-title">GitHub - castacks/tartanvo: TartanVO: A Generalizable Learning-based VO</p><p class="og-desc">TartanVO: A Generalizable Learning-based VO. Contribute to castacks/tartanvo development by creating an account on GitHub.</p><p class="og-host">github.com</p></div></a></div><p><a href="https://youtu.be/NQ1UEh3thbU?si=7MaT-hHiEqLThkjm" target="_blank" class="ke-link">https://youtu.be/NQ1UEh3thbU?si=7MaT-hHiEqLThkjm</a></p><div class="figure-video" data-ke-type="video" data-video-url="https://www.youtube.com/watch?v=NQ1UEh3thbU" data-video-host="youtube" data-video-play-service="daum_cafe" data-video-width="480" data-video-height="360" data-ke-style="alignCenter" data-video-thumbnail="https://scrap.kakaocdn.net/dn/5qC4T/hyZOXBNrFX/TN91CtkizhN2iZItWaZMK1/img.jpg?width=480&height=360&face=0_0_480_360,https://scrap.kakaocdn.net/dn/cEP1iG/hyZONZ5c1G/pGfJej5SoafljdIzXGw0e0/img.jpg?width=480&height=360&face=0_0_480_360,https://scrap.kakaocdn.net/dn/cj5JbN/hyZO0kZfMQ/HIufN2nj4TQtBhvmITKiik/img.jpg?width=480&height=360&face=0_0_480_360" data-ke-mobile-mobilestyle="widthContent" data-video-origin-width="480" data-video-origin-height="360"><iframe src="https://www.youtube.com/embed/NQ1UEh3thbU" width="480" height="360" frameborder="0" scrolling="no" allowfullscreen=""></iframe><div class="figcaption"></div></div><p> </p><p> </p><p> </p><p>Neural Rendering 방식: NeRF-SLAM, Gaussian Splatting SLAM</p><p>>>> 지도를 Point cloud 형식이 아닌 AI 모델 자체로 저장해서 사진처럼 렌더링 하는 방식</p><p> </p><p><b> Hybrid 방식: SuperPoint + SuperGlue</b></p><p>>>> SLAM 전체를 AI로 하진 않고, <b>특징점 찾기만 AI가</b> 하고 나머지는 수학으로 연산함</p><p> </p><p><a href="https://youtu.be/dFXWUuATZT8?si=HJefATAB7ae4Mbpz" target="_blank" class="ke-link">https://youtu.be/dFXWUuATZT8?si=HJefATAB7ae4Mbpz</a></p><div class="figure-video" data-ke-type="video" data-video-url="https://www.youtube.com/watch?v=dFXWUuATZT8" data-video-host="youtube" data-video-play-service="daum_cafe" data-video-width="853" data-video-height="480" data-ke-style="alignCenter" data-video-thumbnail="https://scrap.kakaocdn.net/dn/cDS1IN/hyZNKjDlor/bHb9N7kI4m4BVZ6NCsk8GK/img.jpg?width=1280&height=720&face=0_0_1280_720,https://scrap.kakaocdn.net/dn/rLn1M/hyZOX9D2iM/8oCDuBZKu5OpOtyIVwxCO0/img.jpg?width=1280&height=720&face=0_0_1280_720" data-ke-mobile-mobilestyle="widthContent" data-video-origin-width="853" data-video-origin-height="480"><iframe src="https://www.youtube.com/embed/dFXWUuATZT8" width="853" height="480" frameborder="0" scrolling="no" allowfullscreen=""></iframe><div class="figcaption"></div></div><p><a href="https://github.com/magicleap/SuperGluePretrainedNetwork" target="_blank" class="ke-link">https://github.com/magicleap/SuperGluePretrainedNetwork</a></p><div class="figure-open" contenteditable="false" data-ke-type="opengraph" data-ke-align="alignCenter" data-og-type="website" data-og-title="GitHub - magicleap/SuperGluePretrainedNetwork: SuperGlue: Learning Feature Matching with Graph Neural Networks (CVPR 2020, Oral)" data-og-description="SuperGlue: Learning Feature Matching with Graph Neural Networks (CVPR 2020, Oral) - magicleap/SuperGluePretrainedNetwork" data-og-host="github.com" data-og-source-url="https://github.com/magicleap/SuperGluePretrainedNetwork" data-og-url="https://github.com/magicleap/SuperGluePretrainedNetwork" data-og-image="https://scrap.kakaocdn.net/dn/bqZsGT/hyZOSgbJvD/iKUgtWQZDM4vbkZpkFIGvk/img.png?width=1200&height=600&face=0_0_1200_600,https://scrap.kakaocdn.net/dn/qKpKW/hyZOQCH8xi/L0dyv1XG6Wv218GsXFVhm1/img.png?width=1200&height=600&face=0_0_1200_600"><a href="https://github.com/magicleap/SuperGluePretrainedNetwork" target="_blank" data-source-url="https://github.com/magicleap/SuperGluePretrainedNetwork"><div class="og-image"><img class="thumb_img" src="https://scrap.kakaocdn.net/dn/bqZsGT/hyZOSgbJvD/iKUgtWQZDM4vbkZpkFIGvk/img.png?width=1200&height=600&face=0_0_1200_600,https://scrap.kakaocdn.net/dn/qKpKW/hyZOQCH8xi/L0dyv1XG6Wv218GsXFVhm1/img.png?width=1200&height=600&face=0_0_1200_600" alt="" xxxxonerror="this.src="//img1.kakaocdn.net/thumb/C200x200/?fname=https%3A%2F%2Ft1.daumcdn.net%2Fcafe_image%2Fcafe_meta_image_190529.png""></div><div class="og-text"><p class="og-title">GitHub - magicleap/SuperGluePretrainedNetwork: SuperGlue: Learning Feature Matching with Graph Neural Networks (CVPR 2020, Oral)</p><p class="og-desc">SuperGlue: Learning Feature Matching with Graph Neural Networks (CVPR 2020, Oral) - magicleap/SuperGluePretrainedNetwork</p><p class="og-host">github.com</p></div></a></div><p> </p><p> </p><hr data-ke-style="style6"><p> </p><p> </p><p><a href="https://gazebosim.org/docs/all/ros_installation/" target="_blank" class="ke-link">https://gazebosim.org/docs/all/ros_installation/</a></p><div class="figure-open" contenteditable="false" data-ke-type="opengraph" data-ke-align="alignCenter" data-og-type="website" data-og-title="Installing Gazebo with ROS — Gazebo jetty documentation" data-og-description="What is the difference between a ROS vendor package and other packages listed in this page? The ROS vendor packages are (indeed) ROS packages (built by the ROS buildfarm and hosted in the ROS repository) that provide the installation of the Gazebo librarie" data-og-host="gazebosim.org" data-og-source-url="https://gazebosim.org/docs/all/ros_installation/" data-og-url="https://gazebosim.org/docs/all/ros_installation/" data-og-image="//img1.kakaocdn.net/thumb/C200x200/?fname=https%3A%2F%2Ft1.daumcdn.net%2Fcafe_image%2Fcafe_meta_image_190529.png"><a href="https://gazebosim.org/docs/all/ros_installation/" target="_blank" data-source-url="https://gazebosim.org/docs/all/ros_installation/"><div class="og-image"><img class="thumb_img" src="//img1.kakaocdn.net/thumb/C200x200/?fname=https%3A%2F%2Ft1.daumcdn.net%2Fcafe_image%2Fcafe_meta_image_190529.png" alt="" xxxxonerror="this.src="//img1.kakaocdn.net/thumb/C200x200/?fname=https%3A%2F%2Ft1.daumcdn.net%2Fcafe_image%2Fcafe_meta_image_190529.png""></div><div class="og-text"><p class="og-title">Installing Gazebo with ROS — Gazebo jetty documentation</p><p class="og-desc">What is the difference between a ROS vendor package and other packages listed in this page? The ROS vendor packages are (indeed) ROS packages (built by the ROS buildfarm and hosted in the ROS repository) that provide the installation of the Gazebo librarie</p><p class="og-host">gazebosim.org</p></div></a></div><p><a href="https://community.stereolabs.com/t/stereo-labs-zed-2-simulation-gazebo-plugin/428/3" target="_blank" class="ke-link">https://community.stereolabs.com/t/stereo-labs-zed-2-simulation-gazebo-plugin/428/3</a></p><p> </p><p> </p><p> </p><p> </p>
<!-- -->
첫댓글 가제보시뮬레이터에서 zed 카메라모델 지원하는지 확인할것
가제보 시뮬레이터 설치방법조사, ros2에 통합되어 있는지? 아니면 따로 설치해야 하는지?
방학과제 제안 : RTAB-Map으로 slam, localization한결과와 ros2에 들어있는 2d slam(cartographer), 2d localization(amcl) 의 성능비교 -> RTAB-Map localization결과를 nav2에서 어떻게 사용할건지, 3차원->2차원변환 가능한가?
주행은 2차원 맵과 2차원 좌표로 변환후 하는거라 별차이가 없을것 같지만 nav2로 주행성능까지 비교
실험환경
1. ros2 jazzy 버전사용, unbuntu24.04사용
2. 가제보 시뮬레이터에서 모든 실험 진행 -> 요즘은 ros2에 가제보가 통합되어 있는것 같음
3. 실제 실험시는 jetson orin이나 라즈베리파이5 이용