Target Identification Performance Research Articles

Soil properties and performance of landmine detection by metal detector and ground-penetrating radar — Soil characterisation and its verification by a field test

Metal detectors have commonly been used for landmine detection, and ground-penetrating radar (GPR) is about to be deployed for this purpose. These devices are influenced by the magnetic and electric properties of soil, since both employ electromagnetic techniques. Various soil properties and their spatial distributions were measured and determined with geophysical methods in four soil types where a test of metal detectors and GPR systems took place. By analysing the soil properties, these four soils were classified based on the expected influence of each detection technique and predicted soil difficulty. This classification was compared to the detection performance of the detectors and a clear correlation between the predicted soil difficulty and performance was observed. The detection performance of the metal detector and target identification performance of the GPR systems degraded in soils that were expected to be problematic. Therefore, this study demonstrated that the metal detector and GPR performance for landmine detection can be assessed qualitatively by geophysical analyses.

The crossmodal facilitation of visual object representations by sound: Evidence from the backward masking paradigm.

We report a series of experiments designed to demonstrate that the presentation of a sound can facilitate the identification of a concomitantly presented visual target letter in the backward masking paradigm. Two visual letters, serving as the target and its mask, were presented successively at various interstimulus intervals (ISIs). The results demonstrate that the crossmodal facilitation of participants' visual identification performance elicited by the presentation of a simultaneous sound occurs over a very narrow range of ISIs. This critical time-window lies just beyond the interval needed for participants to differentiate the target and mask as constituting two distinct perceptual events (Experiment 1) and can be dissociated from any facilitation elicited by making the visual target physically brighter (Experiment 2). When the sound is presented at the same time as the mask, a facilitatory, rather than an inhibitory effect on visual target identification performance is still observed (Experiment 3). We further demonstrate that the crossmodal facilitation of the visual target by the sound depends on the establishment of a reliable temporally coincident relationship between the two stimuli (Experiment 4); however, by contrast, spatial coincidence is not necessary (Experiment 5). We suggest that when visual and auditory stimuli are always presented synchronously, a better-consolidated object representation is likely to be constructed (than that resulting from unimodal visual stimulation).

Crossmodal facilitation of masked visual target identification

In the present study, participants identified the location of a visual target presented in a rapidly masked, changing sequence of visual distractors. In Experiment 1, we examined performance when a high tone, embedded in a sequence of low tones, was presented in synchrony with the visual target and observed that the high tone improved visual target identification, relative to a condition in which a low tone was synchronized with the visual target, thus replicating Vroomen and de Gelder's (2000, Experiment 1) findings. In subsequent experiments, we presented a single visual, auditory, vibrotactile, or combined audiotactile cue with the visual target and found similar improvements in participants' performance regardless of cue type. These results suggest that crossmodal perceptual organization may account for only a part of the improvement in participants' visual target identification performance reported in Vroomen and de Gelder's original study. Moreover, in contrast with many previous crossmodal cuing studies, our results also suggest that visual cues can enhance visual target identification performance. Alternative accounts for these results are discussed in terms of enhanced saliency, the presence of a temporal marker, and attentional capture by oddball stimuli as potential explanations for the observed performance benefits.

Target identification and navigation performance modeling of a passive millimeter wave imager

Human task performance using a passive interferometric millimeter wave imaging sensor is modeled using a task performance modeling approach developed by the U.S. Army Night Vision and Electronic Sensors Directorate. The techniques used are illustrated for an imaging system composed of an interferometric antenna array, optical upconversion, and image formation using a shortwave infrared focal plane array. Two tasks, target identification and pilotage, are modeled. The effects of sparse antenna arrays on task performance are considered. Applications of this model include system trade studies for concealed weapon identification, navigation in fog, and brownout conditions.

A Semi-Supervised SAR Target Recognition Based on Laplacian Regularized Least Squares Classification

提出了一种基于核主成分分析(kernel principal component analysis,简称KPCA)和拉普拉斯正则化最小二乘(Laplacian regularized least squares,简称LapRLS)的合成孔径雷达(synthetic aperture radar,简称SAR)目标识别方法.KPCA特征提取方法不仅能够提取目标主要特征,而且有效地降低了特征维数.Laplacian正则化最小二乘分类是一种半监督学习方法,将训练集样本作为有标识样本,测试集样本作为无标识样本,在学习过程中将测试集样本包含进来以获得更高的识别率.在MSTAR实测SAR地面目标数据上进行实验,结果表明,该方法具有较高的识别率,并对目标角度间隔具有鲁棒性.与模板匹配法、支撑矢量机以及正则化最小二乘监督学习方法相比,具有更高的SAR目标识别正确率.此外,还通过实验分析了不同情况下有标识样本数目对目标识别性能的影响.;A Synthetic Aperture Radar (SAR) target recognition approach based on KPCA (kernel principal component analysis) and Laplacian regularized least squares classification is proposed. KPCA feature extraction method can not only extract the main characteristics of target, but also reduce the input dimension effectively. Laplacian regularized least squares classification is a semi-supervised learning method. In the target recognition process, training set is treated as labeled samples and test set as unlabeled samples. Since the test samples are considered in the learning process, high recognition accuracy is obtained. Experimental results on MSTAR (moving and stationary target acquisition and recognition) SAR datasets show its good performance and robustness to azimuth interval. Compared with template matching, support vector machine and regularized least squares learning method, the proposed method gets more SAR target recognition accuracy. In addition, the effect of the number of labeled points on target identification performance is analyzed at different conditions.

HRR Profile을 이용한 이동 표적에 대한 표적 식별 성능 분석

HRR(High Resolution Range) profile은 표적의 고유한 전자기학적 산란 특성을 1차원 레이더 영상으로 보여줄 뿐만 아니라, 잡음에 강인하면서 실시간으로 획득 가능하기 때문에 비행 표적을 식별하기 위한 레이더 신호로 많이 이용되고 있다. 하지만 HRR profile은 레이더와의 상대적인 각도에 대해 민감하게 변하기 때문에 하나의 표적에 대한 훈련 데이터베이스를 구축하려면 엄청난 양의 HRR profile이 필요하다. 이러한 문제를 해결하기 위하여 본 논문에서는 표적의 비행 기동 시나리오에 따른 훈련 데이터베이스 구축 방법을 기술한다. 이 방법을 사용하면 적은 양의 훈련 데이터베이스로도 높은 식별 성능을 얻을 수 있다. HRR(High Resolution Range) profiles show one-dimensional radar images including electromagnetic scattering phenomena of a target. Thus, they are not only robust to noise, but also easily obtainable in a real-time. However, in order to construct a training database for the success of radar target identification, a huge amount of HRR profiles are needed because HRR profiles are highly dependent on the relative angle between the radar and the target. In order to alleviate this difficulty, a database construction method based on the scenarios of target's movement is proposed. The proposed method is able to provide a reliable target identification performance even with a small amount of training database.

Sampling Procedures for Resonance Based Radar Target Identification

The performance of radar target identification using the natural E-Pulse technique is significantly degraded if the sampling rates of the target response and the E-Pulse filter are ignored. In this work, improved sampling procedures over that previously reported by Antony and Shuley (Electron. Lett., vol. 40) in 2004 are proposed. Numerical results demonstrate that the proposed procedures significantly improve the target identification performance.

UNDERCOMPLETE DICTIONARY-BASED FEATURE EXTRACTION FOR RADAR TARGET IDENTIFICATION

Feature extraction is a challenging problem in radar target identification. In this paper we attempt to exploit the sparse property of the scattering signature with a undercomplete dictionary for target identification, and establish a feature extraction scheme based on the undercomplete dictionary. Furthermore, as an application, we present a feature vector, named as the atom dictionary feature, which is extracted from the scattering signatures over a wide-angle sector. Numerical simulation results show that the proposed atom dictionary feature can improve the performance of radar target identification due to the exploitation of the sparse property of the scattering signature.

Object substitution masking: When does mask preview work?

When a target is enclosed by a 4-dot mask that persists after the target disappears, target identification is worse than it is when the mask terminates with the target. This masking effect is attributed to object substitution masking (OSM). Previewing the mask, however, attenuates OSM. This study investigated specific conditions under which mask preview was, or was not, effective in attenuating masking. In Experiment 1, the interstimulus interval (ISI) between previewed mask offset and target presentation was manipulated. The basic preview effect was replicated; neither ISI nor preview duration influenced target identification performance. In Experiment 2, mask configurations were manipulated. When the mask configuration at preview matched that at target presentation, the preview effect was replicated. New evidence of ineffective mask preview was found: When the two configurations did not match, performance declined. Yet, when the ISI between previewed mask offset and target presentation was removed such that the mask underwent apparent motion, preview was effective despite the configuration mismatch. An interpretation based on object representations provides an excellent account of these data.

Relative influences of adjunctive topiramate and adjunctive lamotrigine on scanning and the effective field of view

A subsample of 67 adult patients with partial seizures participating in a randomized, double-blind study comparing the cognitive effects of adjunctive lamotrigine (LTG) and adjunctive topiramate (TPM) was administered Performance On-Line (POL) in addition to a battery of neuropsychological tests at baseline, week 8 and week 16 of treatment. The POL is a self-administered computer task that measures scanning, divided-attention, and the effective field of view. Although the POL does not measure driving performance, POL scores are correlated with driving performance. The results show that adjunctive TPM, but not adjunctive LTG, negatively impacted cognition. Both simple target identification and divided-attention performance on POL were compromised in the TPM group but not in the LTG group. The relative POL impairment associated with chronic TPM treatment was similar to that observed with the acute effects of alcohol with a breath level of .045% or a low dose of alprazolam (0.5mg). Thus, driving-related visual and cognitive skills were compromised by adjunctive TPM treatment. Therapeutic doses of adjunctive TPM pose a potential risk of impaired scanning and divided-attention skills.

A Neural Network Solution for Identification and Classification of Cylindrical Targets above Perfectly Conducting Flat Surfaces

This paper evaluates the radar target identification and classification performance of neural networks. A set of features are derived from scattered fields calculated by using the image technique formulation and Moment Method (MoM). An RBF (Radial Basis Function) networkthat utilizes the feature set is proposed for target identification and classification. The database contains a finite number of samples of cylindrical targets at certain angles. A portion of the database is used to train the networkand the rest is used to test the performance of the neural networkfor target identification and classification. This workaims to find the heights measured from the surface and radiuses of the targets for identification of targets and determine the right target for classification of targets from the scattered field values.

Feature binding and episodic retrieval in blindness for congruent stimuli: evidence from analyses of sequential congruency

Targets are identified more poorly when presented during a congruent cued response than during an incongruent cued response (blindness effect). The authors investigated sequential trial dependencies in the blindness effect. The results show that the size of the blindness effect depends both on the previous cued response-target congruency relationship and on repetition of events from the preceding trial. This finding suggests that cued responses and targets become linked together in a single episodic trace; repeating one of these events from the preceding trial activates the other. Depending on whether the activated representation matches or conflicts with events on the current trial, target identification performance is either facilitated or impaired. Implications for action planning and feature binding are discussed.

Driver landmark and traffic sign identification in early Alzheimer’s disease

Objective: To assess visual search and recognition of roadside targets and safety errors during a landmark and traffic sign identification task in drivers with Alzheimer’s disease. Methods: 33 drivers with...

P-33: Identification of Simulated Targets as a Function of Target and Background Blur

Simulated imagery was used to determine the effect of target and background blur on target identification performance. An interaction between these two variables was found, indicating that greater image detail may not improve performance, and hence may not be required in applications for which high-resolution databases are not readily available.

Driver Identification of Landmarks and Traffic Signs after a Stroke

A study was done to assess the ability for visual search and recognition of roadside targets and safety errors during a landmark and traffic sign identification task in drivers with stroke, that is, drivers who have had a stroke. Visual search for roadside targets during automobile driving can compete for a driver's cognitive resources and may impair driving, especially in drivers with cognitive impairment caused by stroke. Thirty-two drivers with stroke and 137 neurologically normal older adults underwent a battery of visual, cognitive, and motor tests and were asked to report sightings of specific landmarks and traffic signs along a segment of an experimental drive. The drivers with stroke identified significantly fewer landmarks and traffic signs and showed a tendency to make more at-fault safety errors during the task than did control subjects. Roadside target identification performance and safety errors were predicted by scores on standardized tests of visual, cognitive, and motor function. Drivers with stroke are impaired in a task of visual search and recognition of roadside targets whose demands on visual perception, attention, executive functions, and memory probably increased the cognitive load and worsened their driving safety.

Driver Identification of Landmarks and Traffic Signs After a Stroke

A study was done to assess the ability for visual search and recognition of roadside targets and safety errors during a landmark and traffic sign identification task in drivers with stroke, that is, drivers who have had a stroke. Visual search for roadside targets during automobile driving can compete for a driver's cognitive resources and may impair driving, especially in drivers with cognitive impairment caused by stroke. Thirty-two drivers with stroke and 137 neurologically normal older adults underwent a battery of visual, cognitive, and motor tests and were asked to report sightings of specific landmarks and traffic signs along a segment of an experimental drive. The drivers with stroke identified significantly fewer landmarks and traffic signs and showed a tendency to make more at-fault safety errors during the task than did control subjects. Roadside target identification performance and safety errors were predicted by scores on standardized tests of visual, cognitive, and motor function. Drivers w...

Fifty-percent probability of identification <inline-formula><math display="inline" overflow="scroll"><mo>(</mo><msub><mrow><mrow><mi>N</mi></mrow></mrow><mrow><mn>5</mn><mn>0</mn></mrow></msub><mo>)</mo></math></inline-formula> comparison for targets in the visible and infrared spectral bands

We describe a comparison of target identification performance between targets in the IR and visible bands. Increasing levels of Gaussian blur are applied to 12 varying viewpoint aspects of 12 targets in both the IR and visible spectra. Results are analyzed as a function of blur level and target size to give identification probability as a function of resolvable cycles on target. Probability curves for the IR and visible results are compared and the 50% probabilities of identification ( N 50 ) are estimated for both cases. The results show that there is a significant difference in the target acquisition characteristics in the visible and IR wavebands.

Impact of speckle on laser range-gated shortwave infrared imaging system target identification performance

The U.S. Army and the U.S. Air Force are investigating laser range-gated shortwave IR (LRG-SWIR) imaging systems for use in target identification. When coupled to an electron-bombarded CCD, the imaging system can obtain high-resolution images at long ranges. Speckle, an image artifact inherent in laser-illuminated imaging systems, results from interference patterns caused by the coherent illumination. Laser speckle degrades target identification performance but can be reduced by averaging successive LRG-SWIR images. This research is a first attempt at quantifying target identification performance degradation associated with laser speckle. The research begins with a laboratory experiment to verify a speckle model that includes power spectral density and intensity probability density functions. An LRG-SWIR sensor simulation is developed that includes coherent illumination resulting in speckle target images. A field demonstration is performed to verify the fidelity of the simulation. The simulation is then applied to the U.S. Army Night Vision and Electronic Sensors Directorate (NVESD) target identification set with various levels of image averaging and blur. Observer performance results are analyzed in terms of target identification probability and the effects of various levels of blur and speckle are characterized.

One-dimension range profile identification of radar targets based on a linear interpolation neural network

One-dimension range profile can reflect the precise geometric structure features of radar targets. The approach is comprehensively used for radar target identification (RTI), however it varies with target posture. This paper presents a novel neural network model—linear interpolation neural network (LINN) to solve the problem. LINN combines the variation information of one-dimension range profile with its invariant feature information. Simulation results show that this method greatly improves the target identification performance of radar systems.

A comparison of algorithms for subsurface target detection and identification using time-domain electromagnetic induction data

The performance of subsurface target identification algorithms using data from time-domain electromagnetic induction (EMI) sensors is investigated. The response of time-domain EMI sensors to the presence of a conducting object may be modeled as a weighted sum of decaying exponential signals. Although the weights associated with each of the modes are dependent on the target/sensor orientation, the decay rates are a function of the target's composition and geometry and therefore are intrinsic to the target. Since the decay rates are not dependent on target/sensor orientation or other unobservable parameters, decay rate estimation has previously been proposed as a viable method for target identification. The performance attained with Bayesian target identification algorithms operating on the entire time-domain signal and decay rate estimates is compared through both numerical simulations and application to experimental data. The decay rate estimates utilized in the numerical simulations are assumed to achieve the Cramer-Rao lower bound (CRLB), which provides a lower bound on the variance of an unbiased parameter estimate. The simulations as well as results obtained with experimental data show that processing the entire time-domain signal provides better target identification and discrimination performance than processing decay rate estimates.