Enhancing Flood Impact Analysis using Interactive Retrieval of Social Media Images
This work addresses the challenge of efficient flood impact analysis for disaster responders by reducing manual inspection of social media images, though it is incremental as it builds on existing retrieval and feedback methods.
The paper tackled the problem of limited data for timely flood analysis by proposing an interactive content-based image retrieval system with relevance feedback to filter relevant social media images, achieving an improvement in precision from 55% to 87% in top 100 results after 5 rounds of feedback.
The analysis of natural disasters such as floods in a timely manner often suffers from limited data due to a coarse distribution of sensors or sensor failures. This limitation could be alleviated by leveraging information contained in images of the event posted on social media platforms, so-called "Volunteered Geographic Information (VGI)". To save the analyst from the need to inspect all images posted online manually, we propose to use content-based image retrieval with the possibility of relevance feedback for retrieving only relevant images of the event to be analyzed. To evaluate this approach, we introduce a new dataset of 3,710 flood images, annotated by domain experts regarding their relevance with respect to three tasks (determining the flooded area, inundation depth, water pollution). We compare several image features and relevance feedback methods on that dataset, mixed with 97,085 distractor images, and are able to improve the precision among the top 100 retrieval results from 55% with the baseline retrieval to 87% after 5 rounds of feedback.