Description:
The Automated Pallor Detector is a web application, with mobile phone accessibility, that allows users to upload personal eye and tongue images to be screened for potential abnormalities. These images can either be pre-stored on the user machine, or sourced through direct upload from a smart phone camera, depending on the way that it is accessed. The uploaded images are stored in a database located on the virtual machine, and are then pulled to be processed by the back-end, prior to being deleted for patient privacy once the process is completed. The application itself consists of a GUI that gives users full control of navigation, allowing them to upload images, and run the image processing back-end at the click of a button. The final classification results from screening are displayed by way of a message, indicating wether the user images are normal or abnormal in nature.
Specifications:
The application front-end and back-end components are written in Java, and the project's web application archive (WAR) file is deployed to a tomcat server that is hosted on a Microsoft Azure virtual machine (VM) for global access through the web. When users visit the application address, they are essentially visiting the virtual machine, which is ported to a tomcat server that is hosting the deployed application. The server is automatically launched the moment the address is visited, and the application is then run.
Windows VM (Microsoft Azure)
- RAM: 7.00 GB
- Processor: Inter(R) Xeon(R) CPU E5-2660 0 @ 2.20GHz
- System type: 64-bit OS, x64-based processor
Apache Tomcat server
- Version 9
Java Runtime Environment (JRE)
- Version 1.8
Libraries:
- OpenCV
- Apache Commons File Upload
Publications:
[1] Roychowdhury, Sohini, et al. "Computer aided detection of anemia-like pallor." Biomedical & Health Informatics (BHI), 2017 IEEE EMBS International Conference on. IEEE, 2017.
[2] Roychowdhury, Sohini, Paul Hage, and Joseph Vasquez. "Azure-Based Smart Monitoring System for Anemia-Like Pallor." Future Internet 9.3 (2017): 39.
Data: