Week 3

This week we mainly finished the printing of the road and the implementation of the auto-tracking function.

Road Following

Datasets

First we worked out some bugs in the python code. then we printed the roads and made a dataset with 200 road photos.

Training

The completed dataset was imported into the ResNet-18 neural network model for regression training. In addition, there are many open source neural network models on GitHub to choose from, which greatly reduces our workload.

Testing

Finally, we tested the auto-tracking function of Jetbot. By adjusting some parameters of the model for the motor, we successfully implemented the auto-tracking function.

__END__