Still trying to solve the issue to get Class from BoundingBox.
There is also another problem of lag, which has been there since the start of the project. There a lag of about 5 seconds in the video. Tried using compressed image, rect image and raw image but lag is still there.
Due to the lag the duckie overshoots and takes the desired action after a few seconds. Although the code is running fine and giving desired outputs.
Asked Brenton for some help regarding the 'Class' issue.
Tuesday(09/06/2020)
Brenton gave a solution that worked. The solution was to use bounding_boxes[0].Class instead of bounding_boxes[0].
I am able to use Class in the code and hence I was able to get the duckie change its speed based upon what class it is detecting.
Reviewed the lane following code, and refactored it for all the variable that I was using.
Also, uploaded the darnket and the lane follower to the git repository.
Wednesday (10/06/2020)
Whole of the project is working and I have achieved all the milestones of the project.
Made the presentation for final demonstration and also recorded all the videos required for demo and for the promotional video.
Made the promotional video using an online video editing tool.
Started with the reflection report.
Thursday (11/06/2020)
Presentation went well. All the milestones according to the learning contract were completed.
Made final editions to the reflection report.
Submitted the final presentation and the reflection report.