.jpg)
Welcome! We are Team ###! 🎵
Final goal of this project is making PCL(Point Cloud Map) with SLAM, and navigating optimized path with pre-generated map. Our motivation comes from all the complex buildings, such as mall, university, and etc. Every time even this moment, people get suffered with inconvenience by wasting time escaping the maze.
So, many companies are trying to service AR navigation. They make very elaborate map with high quality. However, their limitation is caused by their high quality map. They need at least stereo camera, or fish-eye camera. Sometimes, they make accurate map with strong sensors, even RIDAR. So, only few engineers can make the map while need for this service is large enough, so that these limited people can’t meet the requirement.
So, we decide to use VSLAM to reduce the cost of making map. We need only camera. Not stereo or fish-eye, but only monocular camera that is used for cellular phone, like Galaxy and iPhone. All we need is cellular phone. Users can record video and send it to our server, then can make their own navigation. If more users send video of certain building, we can provide higher and higher quality of map.
Our architecture is composed of Map Generation, Path Planning and AR Navigation parts. In the perspective of development, architecture can be designed with development and application parts.
Main function of development part is generating map with VSLAM. We generate internal map with video, and convert point cloud map into grid map. Then, localize and do path planning at the application part. Finally, we can reach the goal of application part, AR navigating.