Demonstrator 6 - Content-Based Image Retrieval

This demonstrator shows the integration of existing Grid-enabled technology for parallelized Content-Based Image Retrieval (CBIR).

It illustrates search effectiveness using sophisticated similarity measures, comprehensive query models, and expressive feature types.

It achieves querying performance through scalable parallel query execution on the Grid, using efficient index structures.

Where to find it: Click here. You can find more information and screenshots about this demonstrator here.

Target Audience: developer communities, decision-makers, end-users.

How to use it:

Step 1
- Query image can be uploaded from a local file
- Alternatively, users may select from pre-selected query images
- Start feature extraction of query image
Step 2
- Inspect query image features
- Switch between different feature granularities
- Parameterize query (weightings and aggregates)
- Start query processing
Step 3
- Inspect query results
- Fetch subsequently ranked results
- Pick alternative query image

Go back to Demonstrators