Object Detection and Localization from RGB-D Data

Brief description of the task

The main goal of the task will be to implement a processing pipeline for object detection and localization in traffic environment from RGB-D data (i.e. color images with per-pixel depth data). There are many detection algorithms for traffic related objects using artificial neural networks. The task will be to find the best performing ones with available source codes, install them on a dedicated PC, and get familiar with them. Afterwards, the task will be to create a program that will compute 3D positions of the detected objects from the provided depth data. The output of the pipeline should be written into a format called OpenDrive.

The developed pipeline should be tested first on a standard database and then on custom data.

Recommended skills

The student(s) will be required to learn these skills in order to accomplish the given tasks. Knowing them beforehand (at least at a basic level) will be advantageous.

• Python
• PyTorch (framework for artificial neural networks in Python)
• OpenCV (for image processing)

Notes

Team of two students can work on the task.