The MOTION project (iMage and lidar fusiOn for vehicle exTerior InspectiON) is an innovative solution designed to enhance vehicle exterior inspection through cutting-edge image processing techniques. Funded under the European Union’s Horizon 2020 program, it integrates advanced algorithms for image stitching, Structure from Motion (SfM), and illumination-invariant image matching to improve the accuracy and efficiency of defect detection in vehicle inspections.
The project developed a Fast Image Stitching Algorithm (FISA) to create seamless panoramic images of vehicles, to facilitate quick inspections. Additionally, the Structure from Motion (SfM) component reconstructs a 3D model of vehicles, aligning 2D images with vehicle coordinates using LiDAR data and cameras.
The image matching framework, incorporating deep learning techniques such as RoMa and YOLOv8, ensures high-precision feature detection despite challenging lighting conditions.
MOTION’s outcomes have been validated through extensive testing, with the system integrated into the ASSIST-IoT ecosystem. A specialized API has been developed for easy system integration. The project also explored commercialization pathways, identifying market opportunities, but prioritized delivering a research-driven solution for vehicle inspections.
Project Τitle: Motion – iMage and lidar fusiOn for vehicle exTerior InspectiON ++
Project Duration: 9 months (08-05-2023 to 08/02/2024)
Project Framework & Funding: Horizon 2020, Open call under the Assist-IoT Project, €60,000
Scientific coordinator (UNIWA): Associate Professor Lazaros Grammatikopoulos
Consortium – Partners: UNIWA (Photogrammetry Research Unit, Department of Surveying and Geoinformatics Engineering)
Research Publications / Results / Patents
The project outcomes include:
- Development of advanced algorithms for Image Stitching (Fast Image Stitching Algorithm – FISA), Image Matching (Image Matching Framework), and 3D Model Reconstruction using Structure from Motion (SfM).
- Utilization of Artificial Intelligence algorithms (RoMa, YOLOv8, SAM) to enhance inspection accuracy.
- Testing and evaluation of the system under real-world conditions.
- Publication in an international scientific journal:
El Saer, A., Grammatikopoulos, L., Sfikas, G., Karras, G., and Petsa, E., 2024. A novel framework for image matching and stitching for moving car inspection under illumination challenges. Sensors, 24(4), p.1083. (link)
