03/18/2018
Hey everyone, Alex here, reporting in with a big surprise: the image analysis team working on image analysis! Here we have Alex Weck (myself) stitching together the images taken from the multiple lenses of the 200 megapixel array camera, and another student, Rachel Lau, working with the air filters to measure particulate matter concentrations. The team also put together an air sensor to attach to a drone using a custom 3D printed case, and used the drone's camera to detect the direction of maximum air pollution. Incorporated in real time, this would allow the drone to fly towards pollution sources and take PM measurements where the concentration is highest. The last part of the image analsysis project included setting up a GoPro camera to take a timelapse video of the skyline over the course of several hours each day, snapping one frame every 60 seconds. The group also has some of the preliminary data results to showcase here, although with just a few data sets included, the relationship isn't as robust as it will be after a few more iterations! Also included is a plot of the average percent error as a result of adding each image to the algorithm with its correspondence between image entropy or RMS intensity and particulate concentration - the evening out of the line indicates the optimal number of pictures where the predicted relationship stabilizes! Keep on the look out for our image analysis team final results, they won't disappoint.