Protostar Labs
Our team – Mihaela Orić, Marko Šprem, and Mihael Španović – participated in the 14th IEEE WHISPERS Conference on Hyperspectral Image and Signal Processing, held in Helsinki in December 2024. During this remarkable event, we had the opportunity to present our work on the topic "Detection of Adulterants in Chili Powder Using Affordable Methods." The feedback we received was extremely valuable, and discussions with industry and academic colleagues allowed us to gain a deeper understanding of current challenges and future development directions in this field.
One of the key aspects of our research, closely related to our SKORPI project, is the development of an innovative multispectral camera prototype. Traditional multispectral cameras typically use a single optical filter per pixel, whereas our approach replaces this model by using five RGB cameras, each equipped with only one optical filter.
As part of the SKORPI project, we are also exploring advanced techniques for spectral-segmented and spatially structured illumination. This approach optimizes the identification of sample characteristics and quality by integrating a combination of RGB and monochrome cameras, various filters, and specialized lighting. Such a setup enables highly precise measurements and the generation of 3D scene information, while compressed sensing techniques reduce the number of required captures for reliable results.
Additionally, our research leverages deep learning models and neural networks for data analysis, enabling flexible spectral and spatial data acquisition. Deep learning methods also facilitate real-time spectral analysis, allowing our project to pioneer new possibilities in precision analysis and contaminant detection.
The conference provided an excellent opportunity to connect with leading experts in the field and exchange ideas and experiences. We are excited about the opportunities that our engagement with the IEEE WHISPERS community will open for future collaborations.
