Active contours: Generalization of the snake mode

Abstract— Segmentation is one of the fundamental issues in the field of image processing and computer vision. Various approaches include differentiating an object in the image as a final goal or for further processing (medical diagnosis, surveillance, 3‐D reconstruction and more). Snakes, a model proposed by Kass, Witkin, and Terzopoulos in 1987, provides an efficient method for segmenting an object through the minimization of its energy. The advantage of snakes is in its ability to use high‐level data given by the algorithm operator, as opposed to other methods such as the Laplace technique. The snakes model inherently imposes strong constraints on a given image in order to successfully segment an object. In this paper, the use of adjustment methods is described, which allow us to generalize the snake model to a wider range of applications. Through the use of pre‐processing techniques, the model's constraints were softened. The main theoretical model and its use in facing a real life image is presented.     

The Interaction between Micrometer-size Particles and Flat Substrates: A Quantitative Study of Jump-to-Contact

The interaction force acting on an individual micrometer-size polystyrene particle near a flat, electrically conducting substrate has been measured by attaching the particle to an atomic force microscope cantilever. From the spatial dependence of the interaction force, the equations of motion governing a particle near the substrate can be determined. These considerations allow a prediction of the jump-to-contact distance of the particle as it approaches the substrate. This distance is measured as a function of particle radius and compared with predictions based on the relevant interaction force models.     

Effect of humidity on nanoparticle-based chemiresistors: a comparison between synthetic and real-world samples

Chemiresistors based on metal monolayer-capped nanoparticles (MCNPs) are promising candidates for fast, inexpensive, and portable tracing of (bio)chemical species in the gas phase. However, the sensitivity of such sensors to humidity is problematic, limiting their reliable and reproducible application in real-world environmental conditions. In this work, we employed a compensation method to explore the effect of humidity on a single MCNP chemiresistor as well as on an array of MCNP sensors used to analyze either synthetic or real-world samples. We show that an array of MCNP chemiresistors is able to precisely detect and estimate subtle concentrations of (mixtures of) volatile organic compounds (VOCs) under variable backgrounds of 2-83% relative humidity (RH) only after humidity compensation. Humidity effects were also tested in two clinical trials aimed at detecting prostate cancer and breast cancer through exhaled breath analysis. Analysis of the results showed improved cancer detection capabilities as a result of RH compensation, though less substantial than the impact of RH compensation on synthetic samples. This outcome is attributed to one - or a combination - of the following effects: (i) the RH variance was smaller in the breath samples than that in the synthetic samples; (ii) the VOC composition in the breath samples is less controlled than the synthetic samples; and (iii) the responses to small polar VOCs and water are not necessarily additive in breath samples. Ultimately, the results presented here could assist the development of a cost-effective, low-power method for widespread detection of VOCs in real-world applications, such as breath analysis, as well as for environmental, security, and food applications.     

The trajectories of secondary electrons in the scanning electron microscope

Three-dimensional simulations of the trajectories of secondary electrons (SE) in the scanning electron microscope have been performed for plenty of real configurations of the specimen chamber, including all its basic components. The primary purpose was to evaluate the collection efficiency of the Everhart-Thornley detector of SE and to reveal fundamental rules for tailoring the set-ups in which efficient signal acquisition can be expected. Intuitive realizations about the easiness of attracting the SEs towards the biased front grid of the detector have shown themselves likely as false, and all grounded objects in the chamber have been proven to influence the spatial distribution of the signal-extracting field. The role of the magnetic field penetrating from inside the objective lens is shown to play an ambiguous role regarding possible support for the signal collection.     

Stacked multichannel autoencoder – an efficient way of learning from synthetic data

Learning from synthetic data has many important applications in case where sufficient amounts of labeled data are not available. Using synthetic data is challenging due to differences in feature distributions between synthetic and actual data, a phenomenon we term synthetic gap. In this paper, we investigate and formalize a general framework – Stacked Multichannel Autoencoder (SMCAE) that enables bridging the synthetic gap and learning from synthetic data more efficiently. In particular, we show that our SMCAE can not only transform and use synthetic data on a challenging face-sketch recognition task, but that it can also help simulate real images which can be used for training classifiers for recognition. Preliminary experiments validate the effectiveness of the proposed framework.     

Printing Nanostructures with a Propelled Anti‐Pinning Ink Droplet

Striving for cheap and robust manufacturing processes has prompted efforts to adapt and extend methods for printed electronics and biotechnology. A new “direct‐write” printing method for patterning nanometeric species in addressable locations has been developed, by means of evaporative deposition from a propelled anti‐pinning ink droplet (PAPID) in a manner analogous to a snail‐trail. Three velocity‐controlled deposition regimes have been identified; each spontaneously produces distinct and well‐defined self‐assembled deposition patterns. Unlike other technologies that rely on overlapping droplets, PAPIDs produce continuous patterns that can be formed on rigid or flexible substrates, even within 3D concave closed shapes, and have the ability to control the thickness gradient along the pattern. This versatile low cost printing method can produce a wide range of unusual electronic systems not attainable by other methods.     

Gels as Energy Dissipation Media for Energetic Materials Desensitization

The hypothesis that gels, which have a variety of dissipation mechanisms in response to external forces can provide efficient energy deflection media protecting sensitive energetic materials from accidental initiation is explored. Entrapment of explosives, with an emphasis on peroxide materials, in gels is demonstrated as a general method to substantially lower explosives sensitivity diverting stimuli from the sensitive materials to the gel. Friction is a more difficult stimulus to dissipate in comparison to impact. The desensitization method developed, is with minimal material manipulations. Practical considerations as mode of application, solubility, and controlling speed of gelation determine that the method of choice is a sol‐gel process using a mixture of three alkylalkoxysilanes, R_nSi(OR′)_4–n, producing a hybrid organic‐inorganic gel enabling complete triacetone triperoxide (TATP) desensitization at 20–25 % v/v concentrations.     

Historical document enhancement using LUT classification

The fast evolution of scanning and computing technologies in recent years has led to the creation of large collections of scanned historical documents. It is almost always the case that these scanned documents suffer from some form of degradation. Large degradations make documents hard to read and substantially deteriorate the performance of automated document processing systems. Enhancement of degraded document images is normally performed assuming global degradation models. When the degradation is large, global degradation models do not perform well. In contrast, we propose to learn local degradation models and use them in enhancing degraded document images. Using a semi-automated enhancement system, we have labeled a subset of the Frieder diaries collection (The diaries of Rabbi Dr. Avraham Abba Frieder. ). This labeled subset was then used to train classifiers based on lookup tables in conjunction with the approximated nearest neighbor algorithm. The resulting algorithm is highly efficient and effective. Experimental evaluation results are provided using the Frieder diaries collection (The diaries of Rabbi Dr. Avraham Abba Frieder. ).