Scene Of Interest Research Articles

Partial-Aliasing Correlation Filters

Correlation filters (CFs) are useful tools for detecting and locating signals or objects within a larger signal or scene of interest. Typically, these filters are designed during the training stage without worrying about how the cross-correlation between a test signal and the designed CF template will be carried out during the testing or use stage. Because of its computational benefits, the Fast Fourier Transform (FFT) algorithm is usually used for performing cross-correlations, leading to circular correlations and aliasing in the resulting correlation outputs. The aliasing effects can be suppressed by zero-padding, but at the expense of using longer FFTs and thus incurring more computational complexity. In this paper, we present a new approach where CFs are designed to explicitly allow partial aliasing at test time (thus allowing the use of shorter FFTs). This approach of allowing aliasing in the cross-correlation output and explicitly taking such partial aliasing into account when designing the CF is diametrically opposite to the conventional CF approaches which try to avoid aliasing effects. We demonstrate through numerical results that these new partial-aliasing correlation filters (PACFs) achieve better recognition performance than conventional CFs when used in block filtering architectures that allow aliasing.

Uplink Scheduling of Visual Sensors: When View Popularity Matters

Decentralized camera sensors capture a 3D scene of interest from multiple perspectives. The captured video signals need to be transmitted to a central station over a shared wireless channel. A collocated server streams the gathered data to a collection of clients interested in experiencing the scene interactively. We design a constrained optimization framework for sharing the transmission bandwidth of the wireless channel across the sensors such that the average video quality over the client population is maximized. We consider scheduling the uplink resources of the wireless channel at the view or packet level. In the first case, the central station partitions the channel capacity across a select subset of views that are transmitted entirely. In the second case, the station coordinates the packet transmissions of every sensor. We formulate exact and approximate algorithms to solve the optimization of interest. We examine their transmission efficiency via simulation experiments that demonstrate considerable gains over the state-of-the-art and study the impact of view popularity.

Change detection and deformation analysis using static and mobile laser scanning

Laser scanning is rapidly evolving as a surveying technique and is not only used to assess the geometrical state of a scene but also to assess changes in that state. Change detection is, however, a challenging application for several reasons. First, laser scanning is not measuring fixed points, such as a total station does, therefore in general, some interpolation or object extraction method is required. Second, errors that are inevitably present when determining the geometric state of a scene of interest in one epoch will add up when comparing the state between epochs. In addition, data volumes are constantly increasing, therefore processing methods should be computationally efficient. This paper reviews recent methodology in the form of a method breakdown, thereby distinguishing methods aiming at pure binary change detection from methods that in addition want to quantify change. In addition, the direction of a change is discussed, notably in connection with the measurement geometry. Also, the reference state is discussed, which can be in the form of a free form surface, or in the form of some idealized mathematical primitive like a plane. The different methods are presented in connection with applications in fields like structural monitoring, geomorphology, urban inventory and forestry, as considered by the original authors.

Synthetic Aperture Radar Correlation Imaging

In synthetic aperture radar (SAR) imaging, a scene of interest is illuminated by electromagnetic waves. The goal is to reconstruct an image of the scene from the measurement of the scattered waves using airborne antenna(s). A new imaging algorithm, known as correlation imaging, is suggested in which an image is formed of the reflectivity function squared. The algorithm uses what is often called the correlation function in place of the standard slow- and fast-time data. The correlation function is found via a simple preprocessing step applied to the collected data. A backprojection algorithm is formulated using microlocal analysis. It is found that range and cross-range resolution are improved compared to standard SAR imaging. In addition, the point-spread function (PSF) of the correlation image-fidelity operator exhibits reduced side lobes in comparison with the standard SAR PSF. We also show that the contribution to the image from clutter displaying volume scattering behavior is reduced when certain pairs of correlated data sets are avoided. Note that by clutter we refer to scatterers on the ground that are not of interest; we do not address the case of propagating through a random (i.e., cluttered) medium.

Early prediction in collective intelligence on video users’ activity

The huge volume of available video content calls for methods that offer insight to the content without necessitating burdensome users’ extra effort or being applicable to specific types or conditions. Based on experimentation on collective users’ interactions on a controlled user-experiment, this work analyses the results collected following the argument that bell-shaped reference patterns are shown to significantly correlate with scenes of interest for each video, as designated by the viewers. Though, in order to ensure the correlation of results to bell-shaped reference patterns, aggregation of a number of users’ interactions is required. In order to overcome such an impediment and adhere to a real-case cold start scenario, we propose a stochastic transformation of the aggregated users’ interaction signal into a space defined by its correlation to the bell-shaped reference patterns that is shown to offer significant amelioration as to the percentage of users’ interaction required in order to achieve comparable results to the original users’ interaction space. Moreover, to ensure further the realistic character of the proposed scenario, given an amount of already collected users’ interaction, the interaction of new users’ is shown to be predictable using neural network time series prediction and modeling methods. The results received indicate increased accuracy on how one can predict the most important scenes from low quantity early data of users’ interactions as well as future interaction of unique users. In practice, the proposed techniques might improve both navigation within videos on the web as well as video search results with personalised video thumbnails.

Jego Wehrmacht, czyli polsko-niemiecka akademia językowa Janusza Rudnickiego w "Książkach. Magazynie do czytania."

The texts published by Janusz Rudnicki in the quarterly Książki. Magazyn do czytania (Agora) present a new and interesting scene of fighting with the use of the exploited but still constantly updated meanings of war. In the author’s discussions of German literature, the writer, delegated to Polish-German commissions, continually articulates – for instance in the form of phrasemes and phrases – well-known war languages. I use them in the paper as a post-colonial concept of Rudnicki’s supplementation of the Polish-German historical matrix, but also as explanation of the prejudice, stereotypes and tensions persisting in the space of neighborhood communication.

Cross-cultural analysis of neuroelectrical cognitive and emotional variables during the appreciation of TV commercials

In this study we investigated the cerebral activity of a group of Eastern people during the observation of a Western and an Eastern version of the same TV commercial advertising a very popular smartphone. By comparing the electroencephalographic (EEG) signals in theta, alpha and heart rate (HR) activity of the population investigated, we estimated and compared the levels of memorization, attention, pleasantness and emotion perceived. Results present and increase of the pleasantness and emotion while watching particular scenes of interest of the Eastern version of the commercial. These findings suggest that this kind of technology is able to track variation of the cerebral * This work was partially supported by a grant of the Italian Minister of Foreign Affairs for the development of the bilateral scientific relation between Italy and China “Neuropredictor”. G. Vecchiato W. Kong A.G. Maglione P. Cherubino A. Trettel F. Babiloni Neuropsychological Trends – 16/2014 http://www.ledonline.it/neuropsychologicaltrends/ 24 activity related to cognitive and emotional processing across TV commercials. Moreover, there is the possibility to investigate frame segments of particular interest for marketers that could be properly adapted according to the cultural context in which the advertising campaign will be promoted.

Aperture undersampling using compressive sensing for synthetic aperture stripmap imaging

Synthetic aperture imaging is a high-resolution imaging technique employed in radar and sonar applications, which construct a large aperture by constantly transmitting pulses while moving along a scene of interest. In order to avoid azimuth image ambiguities, spatial sampling requirements have to be fulfilled along the aperture trajectory. The latter, however, limits the maximum speed and, therefore, the coverage rate of the imaging system. This paper addresses the emerging field of compressive sensing for stripmap synthetic aperture imaging using transceiver as well as single-transmitter and multi-receiver systems so as to overcome the spatial Nyquist criterion. As a consequence, future imaging systems will be able to significantly reduce their mission time due to an increase in coverage rate. We demonstrate the capability of our proposed compressive sensing approach to at least double the maximum sensor speed based on synthetic data and real data examples. Simultaneously, azimuth image ambiguities are successfully suppressed. The real acoustical measurements are obtained by a small-scale ultrasonic synthetic aperture laboratory system.

Elemental analysis of soils using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) and laser-induced breakdown spectroscopy (LIBS) with multivariate discrimination: tape mounting as an alternative to pellets for small forensic transfer specimens.

Elemental analysis of soil is a useful application of both laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) and laser-induced breakdown spectroscopy (LIBS) in geological, agricultural, environmental, archeological, planetary, and forensic sciences. In forensic science, the question to be answered is often whether soil specimens found on objects (e.g., shoes, tires, or tools) originated from the crime scene or other location of interest. Elemental analysis of the soil from the object and the locations of interest results in a characteristic elemental profile of each specimen, consisting of the amount of each element present. Because multiple elements are measured, multivariate statistics can be used to compare the elemental profiles in order to determine whether the specimen from the object is similar to one of the locations of interest. Previous work involved milling and pressing 0.5 g of soil into pellets before analysis using LA-ICP-MS and LIBS. However, forensic examiners prefer techniques that require smaller samples, are less time consuming, and are less destructive, allowing for future analysis by other techniques. An alternative sample introduction method was developed to meet these needs while still providing quantitative results suitable for multivariate comparisons. The tape-mounting method involved deposition of a thin layer of soil onto double-sided adhesive tape. A comparison of tape-mounting and pellet method performance is reported for both LA-ICP-MS and LIBS. Calibration standards and reference materials, prepared using the tape method, were analyzed by LA-ICP-MS and LIBS. As with the pellet method, linear calibration curves were achieved with the tape method, as well as good precision and low bias. Soil specimens from Miami-Dade County were prepared by both the pellet and tape methods and analyzed by LA-ICP-MS and LIBS. Principal components analysis and linear discriminant analysis were applied to the multivariate data. Results from both the tape method and the pellet method were nearly identical, with clear groupings and correct classification rates of >94%.

Compressed Sensing for Millimeter-wave Ground Based SAR/ISAR Imaging

Millimeter-wave (MMW) ground based (GB) synthetic aperture radar (SAR) and inverse SAR (ISAR) imaging are the powerful tools for the detection of foreign object debris (FOD) and concealed objects that requires wide bandwidths and highly frequent samplings in both slow-time and fast-time domains according to Shannon/Nyquist sampling theorem. However, thanks to the compressive sensing (CS) theory GB-SAR/ISAR data can be reconstructed by much fewer random samples than the Nyquist rate. In this paper, the impact of both random frequency sampling and random spatial domain data collection of a SAR/ISAR sensor on reconstruction quality of a scene of interest was studied. To investigate the feasibility of using proposed CS framework, different experiments for various FOD-like and concealed object-like targets were carried out at the Ka and W band frequencies of the MMW. The robustness and effectiveness of the recommend CS-based reconstruction configurations were verified through a comparison among each other by using integrated side lobe ratios (ISLR) of the images.

A sequential framework for image change detection.

We present a sequential framework for change detection. This framework allows us to use multiple images from reference and mission passes of a scene of interest in order to improve detection performance. It includes a change statistic that is easily updated when additional data becomes available. Detection performance using this statistic is predictable when the reference and image data are drawn from known distributions. We verify our performance prediction by simulation. Additionally, we show that detection performance improves with additional measurements on a set of synthetic aperture radar images and a set of visible images with unknown probability distributions.

Identifying Important Segments in Videos: A Collective Intelligence Approach

This work studies collective intelligence behavior of Web users that share and watch video content. Accordingly, it is proposed that the aggregated users' video activity exhibits characteristic patterns. Such patterns may be used in order to infer important video scenes leading thus to collective intelligence concerning the video content. To this end, experimentation is based on users' interactions (e.g., pause, seek/scrub) that have been gathered in a controlled user experiment with information-rich videos. Collective information seeking behavior is then modeled by means of the corresponding probability distribution function. Thus, it is argued that the bell-shaped reference patterns are shown to significantly correlate with predefined scenes of interest for each video, as annotated by the users. In this way, the observed collective intelligence may be used to provide a video-segment detection tool that identifies the importance of video scenes. Accordingly, both a stochastic and a pattern matching approach are applied on the users' interactions information. The results received indicate increased accuracy in identifying the areas selected by users as having high importance information. In practice, the proposed techniques might improve both navigation within videos on the web as well as video search results with personalised video thumbnails.

Exploiting polarization using multi-frequency SAR data and multi-dimensional time-frequency techniques

During recent years, Time-Frequency (TF) techniques have been introduced to characterize scene polarimetric behaviours using PolSAR data. In this paper, we apply a TF decomposition approach to the analysis of PolSAR data with different frequency bands and show some polarimetric TF behaviours of various scenes. The PolSAR data may be decomposed in azimuth direction, range direction only and in both directions. Two statistical descriptor, i.e., polarimetric TF stationarity and coherence indicators, are used to depict the features of the scene backscattering response. Their individual performance is assessed. The polarimetric TF features, extracted from the scenes of interest, can show some special relevant information compared to the original full resolution case. Moreover, with the availability of new multi-frequency full-polarization high resolution F-SAR data, the results of same site with respect to different frequency bands are exploited and compared with each other.

Sparsity and Structure in Hyperspectral Imaging : Sensing, Reconstruction, and Target Detection

Hyperspectral imaging is a powerful technology for remotely inferring the material properties of the objects in a scene of interest. Hyperspectral images consist of spatial maps of light intensity variation across a large number of spectral bands or wavelengths; alternatively, they can be thought of as a measurement of the spectrum of light transmitted or reflected from each spatial location in a scene. Because chemical elements have unique spectral signatures, observing the spectra at a high spatial and spectral resolution provides information about the material properties of the scene with much more accuracy than is possible with conventional three-color images. As a result, hyperspectral imaging is used in a variety of important applications, including remote sensing, astronomical imaging, and fluorescence microscopy.

Neurophysiological tools to investigate consumer's gender differences during the observation of TV commercials.

Neuromarketing is a multidisciplinary field of research whose aim is to investigate the consumers' reaction to advertisements from a neuroscientific perspective. In particular, the neuroscience field is thought to be able to reveal information about consumer preferences which are unobtainable through conventional methods, including submitting questionnaires to large samples of consumers or performing psychological personal or group interviews. In this scenario, we performed an experiment in order to investigate cognitive and emotional changes of cerebral activity evaluated by neurophysiologic indices during the observation of TV commercials. In particular, we recorded the electroencephalographic (EEG), galvanic skin response (GSR), and heart rate (HR) in a group of 28 healthy subjects during the observation of a series of TV advertisements that have been grouped by commercial categories. Comparisons of cerebral indices have been performed to highlight gender differences between commercial categories and scenes of interest of two specific commercials. Findings show how EEG methodologies, along with the measurements of autonomic variables, could be used to obtain hidden information to marketers not obtainable otherwise. Most importantly, it was suggested how these tools could help to analyse the perception of TV advertisements and differentiate their production according to the consumer's gender.

Compressed Sensing of Monostatic and Multistatic SAR

In this letter, we study the impact of compressed data collections from a synthetic aperture radar (SAR) sensor on the reconstruction quality of a scene of interest. Different monostatic and multistatic SAR measurement configurations produce different Fourier sampling patterns. These patterns reflect different spectral and spatial diversity tradeoffs that must be made during task planning. Compressed sensing theory argues that the mutual coherence of the measurement probes is related to the reconstruction performance of sparse domains. With this motivation, we propose a closely related t%-average mutual coherence parameter as a sensing configuration quality parameter and examine its relationship to the reconstruction behavior of various monostatic and ultranarrow-band multistatic configurations. We investigate how this easily computed metric is related to SAR reconstruction quality.

Photon counting compressive depth mapping

We demonstrate a compressed sensing, photon counting lidar system based on the single-pixel camera. Our technique recovers both depth and intensity maps from a single under-sampled set of incoherent, linear projections of a scene of interest at ultra-low light levels around 0.5 picowatts. Only two-dimensional reconstructions are required to image a three-dimensional scene. We demonstrate intensity imaging and depth mapping at 256 × 256 pixel transverse resolution with acquisition times as short as 3 seconds. We also show novelty filtering, reconstructing only the difference between two instances of a scene. Finally, we acquire 32 × 32 pixel real-time video for three-dimensional object tracking at 14 frames-per-second.

Kilometer-range depth imaging at 1550 nm wavelength using an InGaAs/InP single-photon avalanche diode detector

We have used an InGaAs/InP single-photon avalanche diode detector module in conjunction with a time-of-flight depth imager operating at a wavelength of 1550 nm, to acquire centimeter resolution depth images of low signature objects at stand-off distances of up to one kilometer. The scenes of interest were scanned by the transceiver system using pulsed laser illumination with an average optical power of less than 600 µW and per-pixel acquisition times of between 0.5 ms and 20 ms. The fiber-pigtailed InGaAs/InP detector was Peltier-cooled and operated at a temperature of 230 K. This detector was used in electrically gated mode with a single-photon detection efficiency of about 26% at a dark count rate of 16 kilocounts per second. The system's overall instrumental temporal response was 144 ps full width at half maximum. Measurements made in daylight on a number of target types at ranges of 325 m, 910 m, and 4.5 km are presented, along with an analysis of the depth resolution achieved.

Development of a UAV-MMS-Collaborative Aerial-to-Ground Remote Sensing System – A Preparatory Field Validation

This study proposed the development plan of a novel aerial-to-ground remote sensing (AGRS) system for surveying the land scenes of interest. Specifically, the AGRS system is composed by integrating an unmanned aerial vehicle (UAV) imaging system and a mobile mapping system (MMS), onboard whose platform a control station is also added. The UAV-MMS-collaboration can be classified into two modes - loosely and tightly, respectively related to two efficacy levels of the AGRS - fine-scale mapping in general and target investigating in special cases. The latter scenario can be illustrated by the tasks of fast-responses to the time-critical events, e.g., seeking the accessible roads into disaster areas. These all pose challenging issues. To ensure the premise for AGRS development, a field test was carried out in prior to examine the collaborative effect between its two RS-functional modules. Two typical topics were explored, i.e., self-indicated orthorectification of the UAV images and landcover classification based on information fusion. The final positive results have basically validated the feasibility of the development of the AGRS system.

Plenoptic Layer-Based Modeling for Image Based Rendering

Image based rendering is an attractive alternative to model based rendering for generating novel views because of its lower complexity and potential for photo-realistic results. To reduce the number of images necessary for alias-free rendering, some geometric information for the 3D scene is normally necessary. In this paper, we present a fast automatic layer-based method for synthesizing an arbitrary new view of a scene from a set of existing views. Our algorithm takes advantage of the knowledge of the typical structure of multiview data to perform occlusion-aware layer extraction. In addition, the number of depth layers used to approximate the geometry of the scene is chosen based on plenoptic sampling theory with the layers placed non-uniformly to account for the scene distribution. The rendering is achieved using a probabilistic interpolation approach and by extracting the depth layer information on a small number of key images. Numerical results demonstrate that the algorithm is fast and yet is only 0.25 dB away from the ideal performance achieved with the ground-truth knowledge of the 3D geometry of the scene of interest. This indicates that there are measurable benefits from following the predictions of plenoptic theory and that they remain true when translated into a practical system for real world data.