Back to Portfolio

Project "Fingerprint"

Project "Fingerprint"

Technologies

C++ (with Qt, OpenCV)3D ScanningComputer VisionImage Processing3D ModelingTexture Mapping

About the Project

The "Fingerprint" project was my bachelor's thesis at the Regensburg Medical Image Computing (ReMIC) Laboratory of OTH Regensburg, in collaboration with St. Josef Clinic in Regensburg. In this research-oriented project, I developed an innovative application for measuring so-called "asensitive" (numb) areas on the finger through an interactive process with a conventional camera. The work included evaluating various methods for 3D mapping and surface measurement on 3D objects. Based on the resulting insights, I developed a two-part application.

The patient marks the asensitive or numb area with a pen and scans the affected finger with a modern 3D scanner in the hospital. This provided a high-resolution 3D model on which the first area to be measured could be calculated. In the first part of the program, this marked area had to be measured based on the individually created 3D model. In the second step, patients should be enabled to update their measurement at home to monitor their healing progress themselves, without having to perform the elaborate scan on-site. For this, I developed a desktop application with which the patient can photograph their finger with a conventional webcam from different angles. With these images, the texture of the high-resolution model is then updated, and the measurement of the marked area is repeated with the new marking.

After comprehensive practical tests and comparison of the webcam solution with the results of the high-resolution 3D camera, it unfortunately turned out that the error on average was too large to use the software on a large scale. Nevertheless, it showed potential and could supplement the expensive and elaborate scanning process with the 3D scanner with sufficient further development.

Links

ReMIC Lab Regensburg