Normalization and analys of images
| Reference number | |
| Coordinator | Instant Advice AB |
| Funding from Vinnova | SEK 100 000 |
| Project duration | June 2015 - January 2016 |
| Status | Completed |
Purpose and goal
The project has met its objective to develop and verify algorithms for image normalization and analysis. The result of using these algorithms on the captured images has been validated in comparison with the corresponding image analysis performed by an expert.
Expected results and effects
The result shows that we will be able to use algorithms for image normalization and analysis as an important component in our platform for monitoring self-care treatments. However, the platform must contain decision support which in some cases will require some form of human interaction before the results are presented to the person who undergoes the treatment.
Planned approach and implementation
The project was divided into 4 parts. The first three parts, image normalization, the assessment of the criteria for analysis and the change assessment of the condition were performed with the focus on acne vulgaris. A challenge was, inter alia, to collect a descent number of relevant images. In this project, we collected more than 500 facial images, which, inter alia, must be properly handled on the basis of the Personal Data Law. The fourth part, optimizing the implementation from a time and memory efficiency perspective has essentially been made based on theory and past experiences.