04.11.2024 – 10.11.2024
“Most of the important things in the world have been accomplished by people who have kept on trying when there seemed to be no hope at all.”
-Dale Carnegie
Group Summary
Dearly beloved,
this week has been quite calm for several reasons, mainly because we waited for the PCB’s to be delivered. To our delight they arrived in the middle of the week, meaning that we can begin to properly construct and test the project.
Individual Summaries
Eirik (Data):
For me, this week has been more oriented towards real-time systems, due to the project submission in that subject being on the 14th of November. Though I have still been doing some research about the HOG descriptor, mainly the thing I mentioned last week about marking one person at a time.
The OpenCV library that I am using has several different trackers built-in. All of them have slight variations in terms of accuracy and speed, so I need to test and figure out which will work best with our project. In terms of choosing which target to lock on to, there are two obvious options; lock on the first person the HOG descriptor detects as the target, or lock onto the target that creates the biggest rectangle (i.e. whoever is closest to the camera). In practical use I believe locking to the closest person would be the best solution, given that there are often people standing in the background. Once I have all that decided, the principle can easily be implemented in a simple if-else statement(hopefully).
Robin (Data):
No progress has been made this week on my part.
Fredrik (Data):
Last week when I tried to use the start button the code froze. This week I solved that issue, all i had to do was set both threads to daemon threads, which was easy enough. I also updated the looks of the GUI, but I’m having problems with downloading cv2 so I can’t implement the camera feed onto it. But it looks like this now:
Vetle (Electrical):
Kadir (Mechanical):
This week I practiced FEM simulation, and then tested that on some parts of our system, and I have discussed that with Hans who has more experience than I do.
Hans (Mechanical):
This week I am still offshore, but I have been doing FEM simulations training. And have been able to find a 3D printer onboard, So I have been able to print out Kadirs Redrawn Shoot back system. He did a great job on that last week, all of the parts fit together nicely.
