Research

Spectral measurements and spectral imaging

We represent color by using accurate representation that is based on spectral information. This spectral approach generalizes to electromagnetic radiation beyond visible light as well, like ultraviolet light and infrared radiation. By using our laboratory facilities, the spectral information can be measured by using pointwise measurements, line scanning and imaging techniques. Our novel measurement processes enable spectral imaging in 200–2500 nm region, multiangle measurement and accurate fluorescence characterization.

Computational methods for spectral data and image analysis

The advantage of spectral imaging is the large amount of information it captures, which allows for identification of chemical and optical qualities and structures of object. For efficient practical use, the spectral data needs to be post-processed and analyzed with advanced computational methods. Our group develops these methods for spectral image compression, reconstruction and segmentation.

In remote sensing applications, hyperspectral imaging provides several benefits over standard imaging techniques. Acquired data contains information about spectral signatures of objects in the ground, and therefore provides highly useful starting point for analysis and processing. In forest monitoring via airborne measurements, spectral properties of single trees can be analyzed accurately by using hyperspectral techniques.

Applications

The spectral imaging can be used in many applications. We are using spectral imaging in medical applications, in remote sensing and in industrial quality control.

Spectral imaging is highly useful in industrial quality control. This technique can be used to analyze material properties and as a research tool for collecting information for design of simple sensors. As an example, the sensor design could aim for construction of a practical imaging system that can clearly detect undesirable changes in object quality.

In medical applications, spectral imaging can be used for identification and analysis of important features, such as lesions and abnormal changes in tissue. Multispectral imaging can be useful in diagnosis making and screening, for example in the case of diseases that affect the retina. Also, spectral information allows one to determine other useful biomedical quantities, e.g. pigment concentrations and blood oxygen saturation levels.