Hyperspectral imaging (HSI) offers valuable insights into non-carbon material properties and biological label-free processes. However, current HSI systems are often expensive, bulky, and limit accuracy, speed, and adaptability across use cases. This project aims to develop a compact, scalable, cost-effective HSI system using advanced optics and imaging techniques suitable for diverse applications. Conventional microscopes face challenges such as label requirement, short working distance, and limited field of view and resolution, restricting their ability to perform label-free, highly accurate inspections. By integrating the Ptychographic method with the HS microscope, this project will overcome these limitations and enable superior imaging methodology.
On the hardware side, the work will focus on enhancing optical components, custom spectral filter designs, and optimized illumination settings. The candidate will take responsibility for the design, fabrication, and characterization of key elements of the system, with an emphasis on precision metrology to ensure reliability and reproducibility. The system should include a well-designed, extensible interface to support additional standard sensor types, as needed by users.
On the software side, the project includes building embedded processing pipelines that reduce raw data at the sensor level. The candidate will develop AI and machine learning algorithms for spectral mixing and unmixing, anomaly detection, pattern recognition, and efficient data compression. To ensure practical usability, these models will also be optimized to run efficiently on edge hardware, with support for cloud-based data storage and remote access when needed.
Real-world test cases will include semiconductor metrology for deep 3D trench inspection, biomedical breast cancer classification in pathology, and aquaculture water monitoring for microorganism detection. These scenarios provide practical environments to validate the system’s performance. A similar architecture can be adapted for use in multiple environmental monitoring sectors, food safety, and industrial inspection.
The PhD will be comfortable working across the imaging system, including optical design, simulation, embedded computing, and applied AI. The ideal candidate should be comfortable working at the intersection of hardware and software in different sectors and be interested in developing sensing systems that are accurate, efficient, and usable outside the lab.- Who you
are:
- Master’s
degree in Photonics, Electronics, Physics, or Computer Science
- Who we
are:
- A
university research group at imec-VUB in Brussels, with expertise in optical
imaging systems and signal processing
- Responsibilities:
- Design of optical setup
- Optical system analysis and simulation
- Material characterization
- Developing AI/ML
algorithms
- Literature review and preparation of journal
publications