This week, I focused on testing the integrated EKF pipeline with the data collected from the moored weather balloon test. I helped review the logged IMU and visual data and checked how the filtering pipeline behaved when the visual feature extraction became unreliable due to cloud cover.
I found that the EKF pipeline can still process the sensor data, but the visual updates are less useful when the star detection pipeline fails or produces poor measurements. This showed that we need stronger filtering or masking before passing visual measurements into the estimator.
My progress is on schedule. Next week, I will help finish the final report, poster, and video, and include the testing results from the balloon experiment.