Object Detection Carla

This repository aims to provide an object detection system in carla simulation environment. YOLOv3 algorithm is chosen as a detector system to detect and classify pedestriants, vehicles and objects on the road. This algorithm is based on YOLOv3: An Incremental Improvement which originaly implemented in YOLOv3

Requirements

1. python 3.7
2. Opencv 3.4.2
3. numpy

Installation

To start with the implementation

1. Download CARLA simulator (version 0.9.11).

2. Clone the following repository

git clone https://github.com/shayantaherian/Object_Detection_Carla.git

3. Copy Object_Detection.py in CARLA_0.9.11/PythonAPI/examples folder.#

4. Run CarlaU£4.exe to connect with the server.

5. Finally run Object_Detection.py.

Note that to use yolov3 for this task, it is required to download yolo weights from yolov3.weights. Examples of this detection system in carla environment can be seen as follows: