Internship/m.sc. Thesis: Validation and Benchmarking Framework for Power Line Inspection Algorithms

Vedi l’offerta su Go Lavoro: golavoro.it/annuncio-lavoro/7f5xed

The topic: - Validation and Benchmarking Framework for Power Line Inspection Algorithms_ This internship and thesis topic focuses on developing a comprehensive framework for validating and benchmarking perception algorithms used in UAV-based power line inspection. The objective is to establish standardized, reproducible procedures for evaluating the performance of algorithms that process multimodal UAV data, particularly RGB imagery and LiDAR point clouds. The work will involve designing controlled test environments replicating inspection scenarios with mock-up components, variable lighting, and environmental conditions. Synchronized RGB-LiDAR data will be collected to create reference datasets for algorithm evaluation. The framework will define objective metrics and automated analysis tools to assess detection accuracy, localization precision, robustness, and computational efficiency. A modular benchmarking suite will be implemented to allow easy integration and testing of algorithms, including those for object detection, semantic labeling, and data reduction. Results will be validated using ground-truth annotations and precise geometric measurements. **Requirements**:



- You are a university or high school student in a MINT discipline and looking for hands-on experience.
- You have knowledge of programming in C++ or Python and a basic understanding of cameras and LiDAR sensors.
- You work independently, systematically, and with a high level of commitment, creativity, and team spirit.
- You are an EU citizen or have a valid work permit for Italy.
- You are proficient in English; knowledge of Italian or German is an advantage.
- The internship is carried out in presence at our premises in Bolzano. Your task:
Your work will include setting up test environments, collecting and annotating data, developing evaluation tools, and analyzing algorithmic performance. Depending on the internship duration, specific aspects such as metric definition, dataset generation, or benchmarking automation may be prioritized. The topic can be extended into a bachelor’s or master’s thesis depending on your study program.
- Azienda: Fraunhofer Italia Research scarl
- Sede: Bolzano
- Dipartimento: Automation & Mechatronics Engineering

Vedi l’offerta su Go Lavoro: golavoro.it/annuncio-lavoro/7f5xed