Senior Design Projects
ECS193 A/B Winter & Spring 2023
Image and Video Processing for Pandemic Intelligence
| ********** | |
| Zhaodan Kong | |
| Department of Mechanical and Aerospace Engineering |
Project's details
| Image and Video Processing for Pandemic Intelligence | |
| The senior design project will be part of an NSF-funded project “Transdisciplinary Innovation in Predictive Science for Emerging Infectious Disease and Spillover”. Drawing lessons from COVID-19, the main goal of the project is to develop new sensor, data analysis, robotics, and AI to search for “Pathogen X”. The sponsor/client of the project, Dr. Zhaodan Kong, and his team mainly focus on developing enabling robotics and AI technologies to characterize and sample wildlife reservoirs in remotely settings such as bat caves and rainforests. They have developed a sensor package, which consists of a LiDAR, a thermal camera, an IMU, a volatile organic compound (VOC) sampler, and an onboard processor, for such purpose. The sensor package is currently in handheld form but in the future will be put on an unmanned aerial vehicle (UAV). More details about the project and sensor package can be found here: https://www.dropbox.com/s/ohuqvav3r0vl3vf/Pandemic%20Engineering%20Workshop%20%28Zhaodan%27s%20Slides%29.pptx?dl=0 | |
| Data has been collected in our lab and in the future will be collected in some wildlife reservoirs in California first but potentially in Africa later. The dataset will include data collected by all the sensors, LIDAR, thermal camera, and IMU (data from the VOC sampler will not be included in this senior design project). The main object of the project will be to implement SLAM and other data fusion algorithms to build a 3D map of a wildlife reservoir as well as the animals residing in it (assuming that they are not moving). Ideally an interface will also be developed to enable field scientists to interrogate the results. | |
| - Required: Implement of SLAM and other data fusion algorithms to build a 3D map of a wildlife reservoir as well as the animals residing in it based on data collected by our sensor package. - Optional: Development of an interface to enable field scientists to interrogate the results. |
|
| - Knowledge about fundamental 3D geometry (e.g., 3D coordinate transformation) and perception (e.g., how to work with point cloud data). - Familiar with camera calibration procedures (e.g., how to get intrinsic and extrinsic parameters). - Past experience on Visual-Inertial Odometry (VIO) / Lidar-Inertial Odometry (LIO) / SLAM. - Strong C++ programming skill. - It will be good for the students to know how to use ROS but not mandatory |
|
| ********** | |
| 30-60 min every two weeks | |
| Open source project | |
| Attachment | N/A |
| Yes | |
| Team members | Kay Krachenfels Curtis Stofer Nima Bayati Thomas Ke |
| Rex Liu | |
| N/A |