image: The image is a leaf. Each of the small images shows the amount of light reflected by the leaf at different wavelengths in the red region of the visible spectrum (bright areas are regions of high reflectance). These images reveal details that cannot be observed by eye and could be used, for example, to detect early signs of plant disease. This image was produced using an Image Replication Imaging Spectrometer (IRIS). IRIS allows information about the spatial and spectral characteristics of an object to be captured in a single snapshot, thus providing more information about an object than conventional imaging techniques. The images captured by IRIS contain eight replicated images of the same object, each of which is for a separate spectral pass-band. This type of multi-spectral, spatially-resolved information is necessary for the complete characterisation of any material or object which shows variations in colour or appearance over its surface (e.g. wood, stone, cosmetics, textiles, packaging materials, hair and skin). The IRIS system can be tuned to specific pass-bands in order to capture salient information for each particular application. view more
Credit: Agnieszka Bialek, National Physical Laboratory
An NPL imaging scientist Agnieszka Bialek won the Royal Photographic Society's Selwyn Award for her outstanding work in multi spectral imaging.
The Selwyn Award is sponsored by Imaging Science Group of The Royal Photographic Society. It was introduced in 1994 in memory of E W H Selwyn, an eminent photographic scientist. It is awarded once a year to a scientist under the age of 35 who has conducted successful science based research connected with imaging.
Multi spectral leaf
The image is a leaf. Each of the small images shows the amount of light reflected by the leaf at different wavelengths in the red region of the visible spectrum (bright areas are regions of high reflectance).
These images reveal details that cannot be observed by eye and could be used, for example, to detect early signs of plant disease.
This image was produced using an Image Replication Imaging Spectrometer (IRIS). IRIS allows information about the spatial and spectral characteristics of an object to be captured in a single snapshot, thus providing more information about an object than conventional imaging techniques.
The images captured by IRIS contain eight replicated images of the same object, each of which is for a separate spectral pass-band. This type of multi-spectral, spatially-resolved information is necessary for the complete characterisation of any material or object which shows variations in colour or appearance over its surface (e.g. wood, stone, cosmetics, textiles, packaging materials, hair and skin).
The IRIS system can be tuned to specific pass-bands in order to capture salient information for each particular application.
Multi-spectral microscopic beads
The image is a mixture of microscopic fluorescent beads of different colours, mounted on a microscope slide and illuminated using a 488 nm, 40 mW Coherent Sapphire laser.
This is a simultaneous two-colour fluorescence image produced using Förster Resonance Energy Transfer (FRET) and the Image Replication Imaging Spectrometer (IRIS).
FRET is used to measure distances, detect interactions and monitor conformational changes in biomolecules.
Simultaneous imaging in several wavebands is essential to make measurements of a dynamic system, but this is difficult to achieve with conventional imaging techniques. However by combining FRET with the IRIS system, it has been demonstrated that it is possible to make measurements in four wavebands simultaneously, even at the low intensity levels associated with these measurements.
Science behind the images
Agnieszka works in NPL's Optical Radiation team, in the field of image processing research covering anti-counterfeiting image analysis, molecular imaging, human perception and visualisation.
As part of a large EU-funded project (called MONAT), Agnieszka developed new, cutting-edge image analysis and modelling techniques to determine the mathematical relationship between measured physical properties, the nerve stimuli these properties create, and how the brain interprets these stimuli. This allows scientists to use the physical properties of a new material to predict how natural it will seem.
The predictive power of the models proved to be extremely good for new samples with properties similar to those used to develop the models. Tests also showed that the predictive capability was good (reduction of error value of 77%) even for materials with physical properties falling outside the gamut of those used to develop the models.
Agnieszka's work was crucial to the MONAT project achieving its main goal - to demonstrate the ability to make reliable predictions of complex perceptual attributes (such as naturalness) from just measuring the physical properties of a sample alone.