Biophotonic Sensing Group

Multidimensional statistical analysis in combination with Fourier transformation to improve complex HCA and detect flying insects using Dual-band Scheimpflug Lidar.

Insects play a key role in land and freshwater ecosystems such as food sources, pollinators and nutrient recyclers. As insects are declining at alarming rates, accurate and automated insect monitoring is needed to prioritize habitats for conservation. Dual-band entomological Scheimpflug-Lidar technique is a promising candidate method for real time insect monitoring: it allows the detection of thousands of flying insects per day at high temporal and spatial resolutions. The recorded signals contain a plethora of properties which can be assigned to both instrument-, flight headings- and species-specific clues which may improve classification, but currently remain under exploited. We introduce a systematic approach to robust dimensionality reduction of entomological lidar range-time intensity matrices (2D) of insect observations, into time dependent vectors (1D), and scalar values (0D) which encode features related to both the instrument, flight headings and species characteristics. You will work on a data set with hundreds of thousand entomological lidar observation and employ computational methods with as multidimensional statistical modes in time- and frequency domain, fundamental tone estimation and hierarchical cluster analysis in the complex plane. Your methods will take bioacoustics and species classification to the next level beyond power spectra. Successful thesis work can lead to continuation in our research group.

Trajectory estimation of flying insects using Dual-band Scheimpflug Lidar.

Figure 1: Trajectory-related statistical moments. (A–D) Time-range maps of four insect observations (808 nm signal shown). White circles show the mean pCoM(t) p11 with a horizontal bar indicating its time-domain spread t21 Vertical green bars depict pspread(t). Light-gray and dark-gray circles on the horizontal bar represent skewness in pCoM(t) t31 and pspread(t) t32 respectively.(E, F) Schematics of the CMOS sensor’s partial volumes: near observations appear on “Near pixels,” far observations on “Far pixels.”(G) Circle diameter encodes the angular size pspread(t); circles grow if an insect flies toward the receiver and shrink if it flies away.(H–K) Illustrations of negative versus positive skewness of pCoM(t) (flying down vs. up) and pspread(t) (toward vs. away). Arrow colors in (A–D) match the circles in (G) and dots in (H–K).

Our group develops tools for monitoring insects and their flight directions. Multidimensional statistical moments can be used to determine the trajectory which an insect followed while it was flying through a Lidar beam. Fig.6 illustrates how e.g. two skewness values can be used in combination to deduce if an insect was flying up or down as well as towards or away from the receiver. The project includes the development of an experimental setup, calibration of the setup, in vivo image acquisition of flying insects in the field, data analysis and interpretation. You will learn important approaches whether you choose to pursue industrial or research career.

Development of Marine Lidar for Zooplankton Diversity

Our group develop laser diagnostic tools for our biosphere. In this thesis we aim to profile depth and capture oscillations from aquatic microorganisms in multiple wavelengths for the purpose of discriminating distinct species in marine environment. The project comprise optical design, raytracing, CADing and 3D printing, multivariate calibration and in situ tests of a marine lidar. The technique can shed important light on aquatic ecology with applications for aquafarming. A success project can lead to industrial collaboration for commercialization.

Hierarchical Clustering in the Complex Plane. Applications for Classification Insects in situ.

Do you like complex numbers but haven’t found any use for them? Look no further 😀 Our group develop new rapid tools for monitoring insect diversity in situ. We can collect millions of oscillatory signals from free flying insects and are currently developing computational methods to estimate species richness. You will work on a data set with hundreds of thousand entomological lidar observation and employ computational methods with as multidimensional statistical modes in time- and frequency domain, fundamental tone estimation and hierarchical cluster analysis in the complex plane. Your methods will take bioacoustics and species classification to the next level beyond power spectra. Successful thesis work can lead to continuation in our research group.