Spacial Information Research Articles

Spacial Information Improves Modeling Accuracy in Predicting Student Performance

The relationship between spatial heterogeneity and students' academic performance is a widely researched topic. This study extends the concept of spatial heterogeneity into two categories - allocatable (acquired after birth) and ascribed (determined before birth) - using seating and hometown information as variables in Spatial Durbin Models for statistical analysis, identifying seating as the salient feature. Subsequently, through 2,000 iterations of random state testing with undergraduate data from two classes (N=282), our research compares six representation learning methods, including an improved, custom version of attention-based Graphic Convolutional Network, affirming the superior efficacy of our enhanced approach in extracting seating features. By doing so, our study broadens the application scenarios of social comparison theory and group dynamics theory, while also providing a feasible best-practice reference for network-based instructional modeling.

A novel markov random field segmentation method for SAR images

The synthetic aperture radar(SAR) images is difficult to be segmented by traditional Hidden Markow Random Field(HMRF)-based segmentation method, a novel HMRF-based segmentation method for SAR images is proposed in this article. The method utilizes more spacial information of SAR image model by involving the intensity differences between pixels in the potential function. An initial segmentation result is obtained by K-means method. Then accurate segmentation is carried out by the novel MRF-based segmentation method. The result showed that this method is better than the traditional HMRF one both in noise filtering and miss classification ratio.

Efficient Axillary Lymph Node Detection Via Two-stage Spatial-information-fusion-based CNN

Background and objective: Preoperative imaging diagnosis of axillary lymph node (ALN) metastasis is particularly important for breast cancer patients. This paper focuses on developing non-invasive and automatic schemes for accurate localization and classification (metastasis prediction) of ALN via contrast-enhanced computed tomography (CECT) image and deep learning models. Methods: Based on a two-stage strategy, a novel detection neural network is proposed, where the convolutional block attention module is utilized to extract spacial information and the bottleneck feature fusion module is designed for feature fusion in different scales. Results: Owing to the two embedded modules, the proposed convolutional neural network (CNN) model outperforms Faster R-CNN, YOLOv3, and EfficientDet in the sense that the achieved mAP is 0.454, higher than 0.247, 0.335, and 0.329, respectively. In particular, considering the function of classification only, the proposed model reaches the best performance on most indices (accuracy of 0.968, positive predictive value of 0.972, negative predictive value of 0.966, specificity of 0.983), compared with the methods that have been frequently adopted to predict ALN. In addition, the proposed CNN model has the function of locating ALN, which is lacking in existing models. Conclusions: In this paper, a supervised deep learning method is proposed to detect ALN in CECT images. The positive effect of new added modules are verified, and the benefits of spatial information in ALN detection are confirmed. Further, the two subtasks called localization and classification are evaluated separately, where the proposed model achieves the best performance on most indices. The source code mentioned in this article will be released later.

BCAuth: Physical Layer Enhanced Authentication and Attack Tracing for Backscatter Communications

Backscatter communication (BC) enables ultra-low-power communications and allows devices to harvest energy simultaneously. But its practical deployment faces severe security threats caused by its nature of openness and broadcast. Authenticating backscatter devices (BDs) is treated as the first line of defense. However, complex cryptographic approaches are not desirable due to the limited computation capability of BDs. Existing physical layer authentication schemes cannot effectively support BD mobility, multiple attacker identification and attacker location tracing in an integrated way. To tackle these problems, this paper proposes BCAuth, a multi-stage authentication and attack tracing scheme based on the physical spatial information of BDs to realize enhanced BD authentication security for both static and mobile BDs. After initial authentication based on BD identity with its position information registration, preemptive authentication and re-authentication are performed according to spatial correlation of backscattered signal source locations associated with the BD. By exploiting clustering-based analysis on spacial information, BCAuth is capable of determining the number of attackers and localizing their positions. In addition, we propose a reciprocal channel-based method for BD re-authentication with better authentication performance than the clustering-based method for mobile BDs when the BDs is able to measure received signal strength (RSS), which also enables mutual authentication. We theoretically analyze BCAuth security and conduct extensive numerical simulations with various settings to show its desirable performance.

Manifestation of Physis in Vilnius County Drymonymy

    The current research aims to shed some light on the conceptualization of physis1 in the drymonyms from present-day Vilnius County. The research is limited to the analysis of 65 names selected from the corpus of 787 drymonyms currently officially functioning in the territory of Vilnius County. Data were collected from printed (Onomasticon of Lithuania’s Forests (Part I, 1994)), electronic (Electronic Nature Catalogue (www.ezerai21.lt)) sources, and interactive maps (Portal of Spacial Information of Lithuania (www.geoporta.lt)). The study reveals that the domain of physis, in a broad sense, is an abundant motivation source. The drymonyms were divided into two groups, those of clear semantics and motivation and those of obscure semantics and motivation. The drymonyms of clear semantics and motivation conceptualize dominant vegetation species (21), fauna (9), forest type (15), and the peculiarities of the landscape (17). The group of obscure semantics and motivation presents an overview of names, the conceptualization of which is ambiguous due to the polysemy of the motivating appellative (1), the appellative or nominal nature of themotivating concept (5), or even both (3).

Person re-identification with part prediction alignment

The key to success of person re-identification(re-id) is extracting the discriminative person features. Various part-level feature extraction methods are proposed to capture local person features for re-id. A prerequisite of part feature extraction is that each part should be well located. We believe that ID predictions in different parts of the same image should be consistent. Instead of using the external dataset and pose estimator for guiding, we propose a re-id model with Part Prediction Alignment (PPA), which aims at aligning the predicted distributions between each part. Due to the global feature and local feature contains different spacial information, we consider that the combination of two sides will further improve the detection effect. Therefore, in this paper we adopt the teacher–student training strategy based on PPA for global–local feature extraction, and the global feature extraction branch as a teacher to guide the training of local feature branch. Experimental results on Market-1501, DukeMTMC-reID and CUHK03 (including CUHK03_Detected and CUHK03_Labeled) datasets confirm the effectiveness of our proposal, we achieve Rank1 with 92.4%, 85.1%, 65.5%, 69.2% on Market-1501, DukeMTMC-reID, CUHK03_Detected and CUHK03_Labeled, respectively.

The behaviour of commercial broilers in response to a mobile robot

ABSTRACT 1. Modern broiler production in large sheds holding upwards of 50 000 birds uses indoor climate control based on a handful of fixed-location sensors, often well above the bird-occupied zone. Significant deviations from optimal climate conditions for the birds are common, but installing a higher-density grid of fixed sensors is not cost effective. A robotic platform, moving through the flock of birds and collecting detailed spacial information on a wide range of climate parameters at bird level, enables accurate decisions to be made to optimise the climate in large sheds being made in real time. 2. A preliminary study investigated the feasibility of running a mobile robotic platform among a flock of broiler chickens for an entire 6-week cycle. Bird behaviour in response to the robot was studied. 3. In total, 1597 Ross 308 broiler chicks were housed in a room that was set up to replicate a commercial environment. The robot was driven along a fixed route three times a day (Monday–Friday) for the whole cycle under manual control. Behaviour was studied using camera footage. 4. The birds showed very little ‘startled’ behaviour in response to the robot and were quick to fill the area behind the robot as it moved past. Activity levels increased during robot runs but to a lesser extent than during walk-through by a human stockman. The challenges of operating the robot changed as the birds grew, with many individuals coming into physical contact with the front of the robot later in the cycle. Despite this, very few birds refused to move out of the way completely, and mortality and production parameters remained acceptable throughout. 5. This study has shown that running a mobile robot among a flock of broiler chickens is possible between 3 and 37 d in a cycle and causes limited disruption to bird behaviour. Future work needs to repeat this trial on multiple flocks to see whether running a mobile robot among the broilers can be achieved in a commercial setting, with limited disruption to bird behaviour.

Thresholding based on Grey Levels, Gradient Magnitude and Spatial Correlation

Image segmentation gained significant importance in recent years. The goal of segmentation is partitioning an image into distinct regions containing each pixel with similar attributes. Several Image segmentation techniques exist based on thresholding and clustering. Image segmentation based on thresholding is typically doesn’t find any objects and bounds (lines, curves, etc.) in image. To boost the segmentation performance based on thresholding strategies, a unique strategy that integrates the spacial information between pixel’s is designed. The proposed strategy utilizes pixel’s grey level Gradient magnitude and gray level spacial correlation at intervals a part to construct a unique two dimensional bar graph, known as GLGM & GLSC. This technique is valid through segmenting many real world pictures. Experimental results proved this method outperforms several existing Thresholding strategies.

FlexRQC: Model for a Flexible Robot-Driven Quality Control Station

As the investment on a dedicated quality control stations is not desirable for limited production batches. In general, those systems result in very optimised systems and the lack of flexibility since they are designed for an ad-hoc production. To provide a solution for those cases, a new model to design a flexible quality inspection system is proposed. This paper introduces FlexRQC (Flexible Robotic Quality Control) a model for characterising flexible robot-driven quality control stations. FlexRQC is divided into two domains: The Quality Control Station Domain (QCSD) and the Model Under Inspection Domains (MUID). FlexRQC takes advantage of 3D CAD systems to get spacial information on the quality control station and the quality requirement. The flexibility of the model has been successfully tested in two quality control station setups and various solid rigid objects.

Mapping of Irrigation Service Capacity by Remote Sensing in Batang Tampo Irrigation System-Lintau Buo

Batang Tampo Irrigation System (DI Tampo) area 1,437 ha, has 65 irrigation service are in Sub Basin Batang Tampo. Main river in Sub Basin is 25 km and a part of Sub Basin Kuantan-Inderagiri. This area lied on hilly terrain 350-2262 m above sea level. Normaly, cropping intensity of rice farms is 2 to 2,5. But in dry period in downstream area it has 1 or failed because of water stress or dried. Even The rainfield area is not planted in the period of dry season. The area of rice field along the river or irrigation system show that it has better cropping intensity. It should have to know how is irrigation service capacity of Batang Tampo Irrigation System. Technology of Remote Sensing it can be used to evaluate of the capacity of irrigation sevice by using of information data satelite, like Digital Elevation Model (DEM) for topografi data, image, land level and Enhanced Vegetation Index (EVI) and rain (Herdianto dkk., 2010). This data usely combine with Geographic Information System (GIS) for spacial information data. This study aims to map the ability of DI Tampo service by utilizing remote sensing technology with a vegetation index as its main parameter, upstream - downstream variations, variations otf the position rice farms land level (Hilly or Valley) or terrain land level, and distance of rice fields from irrigation water sources. The method to identification of the capacity irrigation service is EVI by remote sensing and survey by the farmers to verivication data. Hydrology data is collected by the authority. Rice field area is identified by GIS and topography data by DEM. EVI value, which is a representation of photosynthetic activity and canopy density of paddy cropped or cropping intensity on paddy field each year. Average EVI data is collect for five years (2012 – 2017, excepted 2015). The result of this research shows: Irrigation service capacity influences by the topography of rice farms, mainly in the upstream. In upstream, some area of rice field was not cropped of paddy (different crop) because of water scarsity and different land level of rice farm (Hilly and river). In the downstream Irrigation service capacity influences by the distance of paddy field from the water resource.

A pooled Object Bank descriptor for image scene classification

Object Bank (OB) is a high-level image representation encoding semantic and spacial information, and has superior performance in scene classification tasks. However, the dimensionality of OB feature is high, which demands massive computation. Existing dimensionality reduction methods for OB are incapable of achieving both high classification accuracy and substantial dimensionality reduction simultaneously. In order to solve this problem, we propose a threshold value filter pooling method to avoid noise accumulation in histogram-pooling and represent more useful information than max-pooling. We also propose a Matthew effect normalization method to highlight the useful information, and thus boost the performance of OB-based image scene classification. Finally, we apply these two methods in a dimensionality reduction framework to simplify OB representation and construct more proper descriptors, and thus achieve both dimensionality reduction and classification accuracy increase. We evaluated our framework on three real-world datasets, namely, event dataset UIUC-Sports, natural scene dataset LabelMe, and mixture dataset 15-Scenes. The classification results demonstrate that our framework not only obtains accuracies similar to or higher than the original OB representation, but also reduces the dimensionality significantly. The computational complexity analysis shows that it can reduce the time complexity of classification. Therefore, our framework can improve OB-based image scene classification through both computational complexity reduction and accuracy increase.

A Sense and Avoid Algorithm using Surround Stereo Vision for Drones

As applications of drone proliferate, it has become increasingly important to equip drones with automatic sense and avoid (SAA) algorithms to address safety and liability concerns. Sense and avoid algorithms can be based upon either active or passive sensing methods. Each of them has advantages when compared to the other but neither is sufficient by itself. Therefore, especially for application such as autonomous navigation where failure could be catastrophic, deploying both passive and active sensors simultaneously and utilizing inputs from them become critical to detect and avoid objects in a reliable way. As part of the solution, in this paper, we present an efficient SAA algorithm based on input from multiple stereo cameras, which can be implemented on a low-cost and low-power embedded processor. In this algorithm, we construct an instantaneous 3D occupancy grid (OG) map at each time instance using the disparity information from the stereo cameras. Then, we filter noise using spacial information, and further filter noise using a probabilistic approach based on temporal information. Using this OG Map, we detect threats to the drone in order to determine the best trajectory for it to reach a destination.

공간정보기술 동향 분석을 통한 스마트사회 네트워크 전략에 관한 연구

Currently, the previous studies on the concept, configuration, and system regarding spacial information policies show the little progress. However, these studies only propose the basic structure system in terms of a policy theory, seeking and suggesting the basic concept to establish spacial information policies. In the shift of a new paradigm from a PC-based paradigm that began in the mid-2000s to the current smart society based on mobile devises, such as smart phones, this study thus reviewed the strategic direction of the propulsion to establish the desired future strategy for the spatial technology policy with spatial information system that can influence the future national competitiveness of Korea. The results of this study is expected to be used to analyze a variety of issues and technological changes on geospatial intelligence at the global level and to support the establishment of a road-map for policies and research & development on domestic geospatial information.

Tucker decomposition-based tensor learning for human action recognition

The spatial information is the important cue for human action recognition. Different from the vector representation, the spatial structure of human action in the still images can be preserved by the tensor representation. This paper proposes a robust human action recognition algorithm by tensor representation and Tucker decomposition. In this method, the still image containing human action is represented by a tensor descriptor (Histograms of Oriented Gradients). This representation preserves the spatial information inside the human action. Based on this representation, the unknown tensor parameter is decomposed according to the Tucker tensor decomposition at first, and then the optimization problems can be solved using the alternative optimization method, where at each iteration, the tensor descriptor is projected along one order and the parameter along the corresponding order can be estimated by solving the Ridge Regression problem. The estimated tensor parameter is more discriminative because of effectively using the spacial information along each order. Experiments are conducted using action images from three publicly available databases. Experimental results demonstrate that our method outperforms other methods.

空间–数字反应联合编码中的注意机制述评

近几年来，SNARC (空间–数字反应联合编码)效应受到广泛关注，其内在机制也得到深入探讨。本文总结数字认知引起空间注意转移和空间注意线索激活不同数字表征两个方面的文献，试从注意的自上而下(top-down)和自下而上(bottom-up)加工理论来分析SNARC效应中的注意机制，最后从认知控制的角度来讨论自下而上的加工和自上而下的加工之间的交互作用——即主动性和被动性两种认知控制的动态平衡。 The SNARC (Spatial Numerical Association of Response Codes) effect, in that people respond faster with the left-hand side to small numbers and with the right-hand side to large numbers, had in-spired intensive studies on spatial-numerical representation and its underlying mechanisms in recent years. Deheane et al. (1993) indicated that there is a mental number line in brain, in which the smaller number located in the left side and the larger number located in the right side. The spatial position of a number is activated automatically when a number was displayed. Attention plays the non-neglectable role in SNARC effect. Top-down and bottom-up driven processing are two important thinking ways. Bottom-up driven theory holds that the spacial characteristics of number are stable. Numerical and spacial information are activated simultaneously, yet top-down driven theory holds that number have no stable spacial characteristics and SNARC effect is output of subjective strategy control according to tasks. Concerning the role of attention in SNARC, majority studies have been focusing on the spatial characteristics of numbers, e.g. whether numbers can elicit attentional shifts automatically. Several studies have found that small numbers shift spacial attention to the left, while large numbers shift spacial attention to the right (Calabria & Rossetti, 2005; Fias et al., 2001; Fischer et al., 2003). In bottom-up driven processing, posterior parietal cortex (PPC) play a key role. But when participants were told to image a right-to-left ruler before response, the SNARC effect would disappeared or even reversed (Ristic et al., 2006). The prevailing view is that number can elicit attentional shifts, but under the mediation of top-down driven processing. However, new emerging evidence contributed to this topic from a completely new angle, e.g. how attentional cues, such as encoding of spatial locations, might activate numerical re-presentations differently. Recently the studies from endogenous and exogenous paradigms in spa- cial attention cues, visuospacial priming cues, the special information in interoreceptor show that special attentional cues could activate numerical representations automatically: Firstly, comparing with endogenous, exogenous cue has more stable influence than endogenous cue on SNARC effect. In early processing, exogenous cue play a main role, then along with the SOAs (cue-target stimulus onset asynchrony) increased, endogenous cue take the advantage (Pan et al., 2009, 2010, 2011); Secondly, visuospacial and numerical have an interaction before selection stage, that is to say, right visuospacial priming cues activated faster response with large number, while left visuospacial priming cue activated fast response with small number in bottom-up driven way (Herrera & Macizo, 2011); In addition, body movement could be taken as special cue too. Head, eye or whole body move to left direction facilitated the generation of small number, while move to right direction facilitated the generation of large number (Hartmann et al., 2012; Loetscher et al., 2010; Loetscher et al., 2008). In terms of attentional cues activate numerical representations, a bottom-up driven processing deed exist in SNARC effect. By presenting literatures from abovementioned two perspectives of number elicit attention shift and attentional cues activate numerical representations, we try to perform a thorough review and analysis on the potential role of attention involved in SNARC via either a top-down- or a bottom-up-driven processing. Moreover, in cognitive control perspective, the interaction of top-down driven and bottom-up driven process- ing is discussed. The mechanism of SNARC effect may be the balance of proactive control and inactive control. Finally, some potential researches for this topic are also presented and discussed.

3차원 지적정보시스템 개발을 위한 통합 공간정보식별자 개발

The study aims to secure an integrated UFID system-based ID system as it defines the three-dimensional land registration record through selection and standardization of objects. By comparing the integrated ID system secured by the study and the objected ID system proposed by Seoul City, the study came up with practicality of the integrated ID system for the three-dimensional land registration record information system. An integrated UFID-based intellectual spacial information which is consisted of 41 figures in total was developed by the study. The study confirmed the practicality of the integrated ID system by comparing it with the objected ID of the three dimensional land registration record information system established by Seoul City.

The influence of processing of tactile roughness on maintenance of spacial information

The influence of processing of tactile roughness on maintenance of spacial information

Bilayer Segmentation of Live Video in Uncontrolled Environments for Background Substitution: An Overview and Main Challenges

Bilayer segmentation of live video in uncontrolled environments is an essential task for home applications in which the original background of the scene must be replaced, as in video chats or traditional videoconference. The main challenge in such conditions is overcome all difficulties in problem-situations (e.g., illumination change, distract events such as element moving in the background and camera shake) that may occur while the video is being captured. This paper presents a survey of segmentation methods for background substitution applications, describes the main concepts and identifies events that may cause errors. Our analysis shows that although robust methods rely on specific devices (multiple cameras or sensors to generate depth maps) which aid the process. In order to achieve the same results using conventional devices (monocular video cameras), most current research relies on energy minimization frameworks, in which temporal and spacial information are probabilistically combined with those of color and contrast.

Efficient characterization of inhomogeneity in contraction strain pattern

Cardiac dyssynchrony often accompanies patients with heart failure (HF) and can lead to an increase in mortality rate. Cardiac resynchronization therapy (CRT) has been shown to provide substantial benefits to the HF population with ventricular dyssynchrony; however, there still exists a group of patients who do not respond to this treatment. In order to better understand patient response to CRT, it is necessary to quantitatively characterize both electrical and mechanical dyssynchrony. The quantification of mechanical dyssynchrony via characterization of contraction strain field inhomogeneity is the focus of this modeling investigation. Raw data from a 3D finite element (FE) model were received from Roy Kerckhoffs et al. and analyzed in MATLAB. The FE model consisted of canine left and right ventricles coupled to a closed circulation with the effects of the pericardium acting as a pressure on the epicardial surface. For each of three simulations (normal synchronous, SYNC, right ventricular apical pacing, RVA, and left ventricular free wall pacing, LVFW) the Gauss point locations and values were used to generate lookup tables (LUTs) with each entry representing a location in the heart. In essence, we employed piecewise cubic interpolation to generate a fine point cloud (LUTs) from a course point cloud (Gauss points). Strain was calculated in the fiber direction and was then displayed in multiple ways to better characterize strain inhomogeneity. By plotting average strain and standard deviation over time, the point of maximum contraction and the point of maximal inhomogeneity were found for each simulation. Strain values were organized into seven strain bins to show operative strain ranges and extent of inhomogeneity throughout the heart wall. In order to visualize strain propagation, magnitude, and inhomogeneity over time, we created 2D area maps displaying strain over the entire cardiac cycle. To visualize spatial strain distribution at the time point of maximum inhomogeneity, a 3D point cloud was created for each simulation, and a CURE index was calculated. We found that both the RVA and LFVW simulations took longer to reach maximum contraction than the SYNC simulation, while also exhibiting larger disparities in strain values during contraction. Strain in the hoop direction was also analyzed and was found to be similar to the fiber strain results. It was found that our method of analyzing contraction strain pattern yielded more detailed spacial and temporal information about fiber strain in the heart over the cardiac cycle than the more conventional CURE index method. We also observed that our method of strain binning aids in visualization of the strain fields, and in particular, the separation of the mass points into separate images associated with each strain bin allows the strain pattern to be explicitly compartmentalized.

SERS imaging for label-free detection of the phospholipids distribution in hybrid lipid membrane

A label-free and low-cost mapping method based on SERS imaging was reported for illustrating the distribution of phospholipids with similar structures in binary lipid membranes on Ag nanoparticles (Ag NPs) films. The Ag NPs films exhibited strong SERS activity and good reproducibility which were investigated with p-aminothiophenol (p-ATP) as probe molecules. Atomic force microscope (AFM) measurement proved that compact lipid membranes formed on the Ag NPs films. Basing on the Ag NPs films, the SERS spectra of phospholipids in the mixed lipid membranes were achieved and the inherent vibration of 1,2-dimyristoyl-sn-lycero-3-phosphoglycerol, sodium salt (DMPG), 1482 cm−1, was used to distinguish between DMPG and dimyristoylphosphatidylcholine (DMPC). The proportions of phospholipids in the mixed lipid membranes were represented by the intensity ratio of peaks at 1482 cm−1 and 1650 cm−1 (R 1482/1650) simultaneously: increasing R 1482/1650 indicated higher proportion of DMPG and lower proportion of DMPC. SERS imaging of the lipid membranes was constructed as a combination of spacial information and the semiquantitative detection of phospholipids according to R 1482/1650, which showed that the charged phospholipids, DMPG, aggregated in the hybrid lipid membranes. The presented mapping strategy based on SERS imaging carried out on Ag NPs films supplied a facile, label-free and inexpensive way for potential applications in the research on the structure of the lipid membrane, such as lipid domains and rafts.