• Development of software in C++ to improve reliability of optical character recognition.• Implementation of automated stitching and corrective orthogonal projection of large scale ancient documents that are piece-wise digitalized.• Direct support contact to our internal customers in the product lines to provide solutions for both short and mid term problems resulting in financial improvements.

Study of the effect of pharmacological drugs (ex: paracetamol) on the brain. In a multi-field team (doctors, biochemists, physicists NMR...), I am responsible for the improvement of the MRI images processing. I learned about the platform and the software used in the laboratory: AIR for the registration of the images and SPM for the detection of neuronal activation. I developed a semi-rigid algorithm of registration as well as a program of segmentation and quantification including its graphic user interface under MatLab (Windows). I digitized several cerebral atlases of: rat, monkey and mouse. I am currently re-implementing algorithms in C++ of image processing within the framework of the migration under a free open source software.

As part of the restoration of monuments, the team of geology of Ghent studies the erosion of the stone and its treatments. The 3D X-ray tomography is a way of analyzing the structure of the stone. I was in charge of the development of the software for the analysis of these images, their segmentation and their morphological characterization. I discussed with the experts to identify their needs and to establish a specification. I developed algorithms under MatLab (Windows) for automatic calculation and analysis of porosity on these 3D images, and I adapted a graphic user interface to it. Then, I tested and validated the whole tool. I wrote the corresponding documentation.

Optical interferometry is used to obtain astronomical images of higher resolution. This method generates an incomplete plan of Fourier and among others things, my team was responsible for the restoration of these images. My task in the team was the study the different a priori needed to complete the data of this ill-posed problem. After having taken knowledge of the specifications, I got familiar with the field of interferometry and astronomy. To integrate new criteria in the existing programs in IDL (Unix), I calculated the derivative of complex criteria (manually and using Maple). I analyzed the results and selected the best criteria.

GE develops and sells numerical mammography’s appliances. With the goal of an optimal use of these new devices, my team worked out a series of quality assurance protocols. My mission was to define the quality protocol of the display chain and to create a test pattern. Using DICOM (Digital Imaging and Communication in Medicine) standards for the protocols of monitors’ calibration and different display’s LUT (Look Up Table) which I learned for the project, I created a pattern morphologically close to a mammography in IDL (Unix).