After the symposium we continued to work on finding a better way to analyze and create a vector map from the raw lidar data. The first approach we tried was to use a software that was able to read the raw data and compress it into a 3D image. We were ultimately unsuccessful in opening the raw data in the program. Next we tried using matlab which surprisingly worked, but could only open one out of the 30 raw data files at once. The following picture is a matlab rendering of the raw data.
After attempting to continue to find a program that could render the raw lidar data as a 3D image we ultimately were unsuccessful. After Dr. Chen Yu and Xin Long went out to collect more GPS data they soon realized that we could actually rely on GPS data from google maps, due to an unforeseen circumstance. For the google maps data to be accurate the position of the GPS in the google maps vehicle need to correlate exactly to the GPS unit in our test vehicle. In this case the GPS unit was mounted in the center of the google maps vehicle and also to the center of the test vehicle. This is not as common due to the placement of the GPS unit, but coincidentally the Lincoln mkz had a GPS unit that was center mounted. This meant that we did not have to deal with the complex raw lidar data. Overall, I enjoyed my experience working with my fellow lab mates at the Berkeley deep drive lab. Initially I found my experience to be more tedious as I had to learn how to code in python, but soon after when I got assigned to the vector map project I found my experience to be more interesting and rewarding as it involved high level of problem solving.
After attempting to continue to find a program that could render the raw lidar data as a 3D image we ultimately were unsuccessful. After Dr. Chen Yu and Xin Long went out to collect more GPS data they soon realized that we could actually rely on GPS data from google maps, due to an unforeseen circumstance. For the google maps data to be accurate the position of the GPS in the google maps vehicle need to correlate exactly to the GPS unit in our test vehicle. In this case the GPS unit was mounted in the center of the google maps vehicle and also to the center of the test vehicle. This is not as common due to the placement of the GPS unit, but coincidentally the Lincoln mkz had a GPS unit that was center mounted. This meant that we did not have to deal with the complex raw lidar data. Overall, I enjoyed my experience working with my fellow lab mates at the Berkeley deep drive lab. Initially I found my experience to be more tedious as I had to learn how to code in python, but soon after when I got assigned to the vector map project I found my experience to be more interesting and rewarding as it involved high level of problem solving.

Comments
Post a Comment