Color normalization makes a Content Based Image Retrieval model more focused on shape and texture features.

Hi, world. This is time to talk about my last secondment at UGR, under supervision of Prof. Rafael Molina and collaborating with Feranado Pérez Bueno.


The secondment commenced in Jun of 2021.

The main goal of this secondment was to analyze the impacts of color normalization on the results of Content-Based Image Retrieval (CBIR) which is completely achieved.

In this project, different medical centers are collaborating to provide a histopathological data set. Each center has its own equipment and material for cutting and staining the tissue. This variety in material brings variety to the color of the images. In the other words, among all important features, the CBIR model would rely on the color features more than others to find similarities. This may cause a problem because the model ranked the images from the other centers with less priority, although they have similarities in the shape features but not in color features. To tackle this problem, color normalization as a preprocessing step is needed to keep the color features of different centers at the same density.

At UGR, I had a great chance to collaborate with Feranado Pérez Bueno, once again, after my secondment in Norway. We had weekly meetings during these 2 months to exchange our knowledge about CBIR and color normalization. In addition to these meetings, I had a great opportunity to have weekly meetings with Prof. Rafael Molina to share the obtained results and have an interactive discussion on possible advancements. As a result of this collaboration, we could start writing an article to report our promising results. If anyone is interested, they can see the published article in my blog post shortly after this one.

Besides all these working meetings, thanks to Dr. Miguel López Pérez, we had some international meetings for language exchange to get to know Chinese, Spanish, English, and Persian and get to know the cultures behind them. One of my goals in my life is to find the best version of my personality. I firmly believe that it cannot be possible without communicating with people with different perspectives on life. Thanks to CLARIFY project we all have this chance to visit new places and learn from new cultures.

Special thanks to prof. Rafael Molina for his warm hospitality.


Language exchange session with Arne schmidt, Jiahui Geng from CLARIFY, and all my friends from China and Spain.

A nice night with beautiful people at UGR. Thanks to Prof. Rafael Molina.


Zahra Tabatabei – ESR9.