We present a study of border surveillance systems for automatic threat estimation. The surveillance systems should allow border control operators to be triggered in time so that adequate responses are possible. Examples of threats are smuggling, possibly by using small vessels, cars or drones, and threats caused by unwanted persons (e.g. terrorists) crossing the border. These threats are revealed by indicators which are often not exact and evidence for these indicators incorporates significant amounts of uncertainty. This study is linked to the European Horizon 2020 project ALFA, which focuses on the detection and threat evaluation of low flying objects near the strait of Gibraltar. Several methods are discussed to fuse the indicators while taking the uncertainty into account, including Fuzzy Reasoning, Bayesian Reasoning, and Dempster-Shafer Theory. In particular the Dempster-Shafer Theory is elaborated since this approach incorporates evaluation of unknown information next to uncertainty. The method is based on belief functions representing the indicators. These functions show a gradual increase or decrease of the suspiciousness depending on input parameters such as object speed, size etc. The fusion methods give two output values for each track: a suspect probability and an uncertainty value. The complete dynamic risk assessment of detected flying objects is evaluated by the automatic system and targets with probabilities exceeding a certain threshold and appropriate uncertainty values are presented to the border control operators.
Particle defects are important contributors to yield loss in semi-conductor manufacturing. Particles need to be detected
and characterized in order to determine and eliminate their root cause. We have conceived a process flow for advanced
defect classification (ADC) that distinguishes three consecutive steps; detection, review and classification. For defect
detection, TNO has developed the Rapid Nano (RN3) particle scanner, which illuminates the sample from nine azimuth
angles. The RN3 is capable of detecting 42 nm Latex Sphere Equivalent (LSE) particles on XXX-flat Silicon wafers. For
each sample, the lower detection limit (LDL) can be verified by an analysis of the speckle signal, which originates from
the surface roughness of the substrate. In detection-mode (RN3.1), the signal from all illumination angles is added. In
review-mode (RN3.9), the signals from all nine arms are recorded individually and analyzed in order to retrieve
additional information on the shape and size of deep sub-wavelength defects. This paper presents experimental and
modelling results on the extraction of shape information from the RN3.9 multi-azimuth signal such as aspect ratio,
skewness, and orientation of test defects.
Both modeling and experimental work confirm that the RN3.9 signal contains detailed defect shape information. After
review by RN3.9, defects are coarsely classified, yielding a purified Defect-of-Interest (DoI) list for further analysis on
slower metrology tools, such as SEM, AFM or HIM, that provide more detailed review data and further classification.
Purifying the DoI list via optical metrology with RN3.9 will make inspection time on slower review tools more efficient.