UFLD Documentation¶
Our project consists of two parts. One part is about generating datasets from CARLA, which is topic of this documentation. The other part is about a neural network, that is able to detect lanemarkings on a road. The documentation about this part can be found here.
Project overview¶
The project is split up into two parts:
The first part covers how to collect labeled images from CARLA.
The second part covers how to generate a dataset from the collected labeled images.
Datasets¶
In order to train a Deep Convolutional Neural Network, we have to create a dataset, containing images and labels, which indicate the lanemarkings on a road. For this purpose, the dataset has to meet some requirements to be read by the Deep Convolutional Neural Network. If the dataset doesn’t fulfill the specification exactly it will fail.
Collecting data from CARLA¶
CARLA is one way to get training and test data. Sadly CARLA doesn’t provide the required information over its API. As a result it’s quite complicated to generate a dataset from CARLA. We wrote some scripts which make this possible. The scripts are located in the scripts folder. If you want to use them correctly, look on this HOWTO which explains their usage in detail.