Fermín A. Colón López

Developed image understanding algorithms to detect printing artifacts and improve printing quality for commercial printers. These innovative algorithms would allow for development of real-time inspection systems to increase profitability by reducing down time. Led team of researchers with the goal to identify future research technologies and presented proposal of research opportunities that became part of our group's technology strategy.

Modeled Image Quality performance of next-generation electro-optical systems. Designed and conducted experiments to predict NIIRS values and proposed extensions and improvements to simulation tools and methodologies. Investigated and recommended a new form of the Image Quality Equation with improved predictability. Co-designed an image chain for modeling advanced EO imaging concepts. Documented modeling scripts to accelerate the training of junior engineers.

Conducted research to develop a new machine learning (ML) framework for printing models. The framework uses a bank of feed-forward neural networks to analyze the digital input and printed output of halftone patterns using a parametric representation of the human visual system texture space and their morphological properties. The model produces realistic dot structures, provides accurate predictions for tone and noise, is easy to implement and computationally fast, and has the potential to be independent of halftone pattern and type of printer. This new framework addresses the limitations of current printer models.
Thesis TItle:
A New Framework For An Electrophotographic Printer Model

Evaluated image quality of inkjet printers and certified new image processing algorithms. Co-developed a tool suite for evaluating and optimizing prototype inkjet printers and developed an algorithm for estimating image quality metrics from natural images. Conducted study on the variability of drop on demand inkjet printing systems and led team to create an algorithm robustness certification process, resulting in improved reliability and robustness of algorithms.

Created image quality test plans, system specifications, and conducted end-to-end system evaluations for digital products for consumer and commercial markets. Modeled and optimized image chains for various digital imaging systems, including KIMES, Disney Fantasy Theater, TRIP, ROVER, and DCL imaging system, among others.

Developed and scale-up laser printing and biotechnology medical films using Design of Experiments (DOE) and Statistical Modeling Techniques (SMT). Co-led development of Infrared (IR) laser printing films and implemented novel formulation changes to improve product quality and reduce waste. Applied these techniques to develop and scale a variety of Health Imaging Products such as DNA sequencing, Dental, and radiographic films. Led team that studied co-development of emulsions, coating, and chemical technologies, establishing guidelines that shortened manufacturing scale-up cycle time and reduced the number of scaling errors.

Demonstrated capability of building novel room light dental films. Coordinated implementation of statistical process control and analysis tools for manufacturing support. Contributed to the development of mixing and reactant delivery technology, enabling precise control of reactants and the development of T-grains, which became the new industry standard. Developed proprietary cleaning methods for insoluble precipitate byproducts in mixing apparatus.

Taught graduate and undergraduate courses in applied Statistics, Research Methods, Applied Mathematics and Physics.