The seventh International Association for Spectral Imaging Meeting will start off with a workshop on Sunday, June 17. The co-instructors Sergey Kucheryavskiy and José Manuel Amigo Rubio will provide a fun, interactive and informative afternoon comparing and contrasting traditional image analysis (digital images) and the state-of-the-art in hyperspectral applications. The objective is to look at the many sides of hyperspectral imaging to provide a better understanding of both approaches, find the potential for synergy between the paradigms of mixing pure spatial and pure spectral approaches and foster new ideas for research and collaborations.

The course will start with an overview of main methods and techniques widely used for analysis of hyperspectral images, including pre-processing of spectra and different ways for producing of spectral maps— from using the spectral intensities at a selected wavelength/waveband to various transformations such as PCA, MCR, clustering, classification, SIMPLISMA etc. Attendees will see, that similar methods do also exist in analysis of traditional images, e.g. transformation between colour modes and models.

The second part will be devoted to the methods HSI community has taken from the traditional image analysis and incorporated to HSI analytical procedures, such as using spatial constrains in MCR. Then we will discuss method which are well applicable to both conventional digital images and spectral maps: image enhancement (contrast, sharpness), detection of edges, detection and analysis of particles, spatial de-noising, etc.

The last part will be devoted to the methods not yet popular for spectral image analysis but quite common in analysis of conventional digital images. This is, first of all, various geometrical transformations, including correction of distortions and interpolation of pixels’ intensity. Another interesting niche is analysis of textures and the use of textural patterns for segmentation of images.

Instructors

Sergey Kucheryavskiy is an associate professor at Department of Chemistry and Bioscience, Aalborg University, Denmark. He has 15+ years of experience in processing and analysis of conventional digital images with applications in remote sensing, medicine, process control and many others.

José Manuel Amigo Rubio is an associate professor at Department of Food Sciences, University of Copenhagen, Denmark. He has 10+ years of experience in hyperspectral imaging and its application in many diverse fields (forensics, pharmaceutical production, food process control, etc.). Moreover, he is the main developer of HYPER-Tools (freely available at https://www.hypertools.org/).