Titre Predicting Changes in Sewer Pipeline Size from Inspection Videos Using Time Series Models

This presentation addresses the problem of predicting changes in sewer pipeline size from inspection videos. We specifically focus on inspection television (ITV) videos of wastewater pipes, which play a crucial role in the management and maintenance of urban networks. On one hand, they help identify anomalies that may affect the pipes, such as obstructions or degradations. On the other hand, they provide essential information about the structural properties of the pipes and networks, including their diameter and the direction of wastewater flow. We propose a classification algorithm for ITV videos, with a particular focus on detecting diameter changes within the pipes. This task is essential for predictive maintenance and hydraulic modeling of wastewater networks. We build on Video Vision Transformer (ViViT)-based methodologies for video classification, which allow for the effective capture of both spatial and temporal relationships between the different images or frames in the video data. We specifically describe different mechanisms for generating training datasets from a subset of manually annotated images. The experimental study shows promising results on real-world ITV video data.