Local Spatial Interaction Research Articles

Ship Detection in SAR Images Based on Feature Enhancement Swin Transformer and Adjacent Feature Fusion

Convolutional neural networks (CNNs) have achieved milestones in object detection of synthetic aperture radar (SAR) images. Recently, vision transformers and their variants have shown great promise in detection tasks. However, ship detection in SAR images remains a substantial challenge because of the characteristics of strong scattering, multi-scale, and complex backgrounds of ship objects in SAR images. This paper proposes an enhancement Swin transformer detection network, named ESTDNet, to complete the ship detection in SAR images to solve the above problems. We adopt the Swin transformer of Cascade-R-CNN (Cascade R-CNN Swin) as a benchmark model in ESTDNet. Based on this, we built two modules in ESTDNet: the feature enhancement Swin transformer (FESwin) module for improving feature extraction capability and the adjacent feature fusion (AFF) module for optimizing feature pyramids. Firstly, the FESwin module is employed as the backbone network, aggregating contextual information about perceptions before and after the Swin transformer model using CNN. It uses single-point channel information interaction as the primary and local spatial information interaction as the secondary for scale fusion based on capturing visual dependence through self-attention, which improves spatial-to-channel feature expression and increases the utilization of ship information from SAR images. Secondly, the AFF module is a weighted selection fusion of each high-level feature in the feature pyramid with its adjacent shallow-level features using learnable adaptive weights, allowing the ship information of SAR images to be focused on the feature maps at more scales and improving the recognition and localization capability for ships in SAR images. Finally, the ablation study conducted on the SSDD dataset validates the effectiveness of the two components proposed in the ESTDNet detector. Moreover, the experiments executed on two public datasets consisting of SSDD and SARShip demonstrate that the ESTDNet detector outperforms the state-of-the-art methods, which provides a new idea for ship detection in SAR images.

VPint: value propagation-based spatial interpolation

Given the common problem of missing data in real-world applications from various fields, such as remote sensing, ecology and meteorology, the interpolation of missing spatial and spatio-temporal data can be of tremendous value. Existing methods for spatial interpolation, most notably Gaussian processes and spatial autoregressive models, tend to suffer from (a) a trade-off between modelling local or global spatial interaction, (b) the assumption there is only one possible path between two points, and (c) the assumption of homogeneity of intermediate locations between points. Addressing these issues, we propose a value propagation-based spatial interpolation method called VPint, inspired by Markov reward processes (MRPs), and introduce two variants thereof: (i) a static discount (SD-MRP) and (ii) a data-driven weight prediction (WP-MRP) variant. Both these interpolation variants operate locally, while implicitly accounting for global spatial relationships in the entire system through recursion. We evaluated our proposed methods by comparing the mean absolute error, root mean squared error, peak signal-to-noise ratio and structural similarity of interpolated grid cells to those of 8 common baselines. Our analysis involved detailed experiments on a synthetic and two real-world datasets, as well as experiments on convergence and scalability. Empirical results demonstrate the competitive advantage of VPint on randomly missing data, where it performed better than baselines in terms of mean absolute error and structural similarity, as well as spatially clustered missing data, where it performed best on 2 out of 3 datasets.

Spatial Interaction Modeling of OD Flow Data: Comparing Geographically Weighted Negative Binomial Regression (GWNBR) and OLS (GWOLSR)

Due to the emergence of new big data technology, mobility data such as flows between origin and destination areas have increasingly become more available, cheaper, and faster. These improvements to data infrastructure have boosted spatial and temporal modeling of OD (origin-destination) flows, which require the consideration of spatial dependence and heterogeneity. Both ordinary least square (OLS) and negative binomial (NB) regression methods have been used extensively to calibrate OD flow models by processing flow data as different types of dependent variables. This paper aims to compare both global and local spatial interaction modeling of OD flows between traditional and geographically weighted OLS (GWOLSR) and NB (GWNBR) modeling methods. From this study with empirical data it is concluded that GWNBR outperforms GWOLSR in reducing spatial autocorrelation and in detecting spatial non-stationarity. Although, it is noted that both local modeling methods show improvement when compared against the equivalent global models.

MRI Image Segmentation Using Conditional Spatial FCM Based on Kernel-Induced Distance Measure

Fuzzy C-means (FCM) clustering is the widest spread clustering approach for medical image segmentation because of its robust characteristics for data classification. But, it does not fully utilize the spatial information and is therefore very sensitive to noise and intensity inhomogeneity in magnetic resonance imaging (MRI). In this paper, we propose a conditional spatial kernel fuzzy C-means (CSKFCM) clustering algorithm to overcome the mentioned problem. The approach consists of two successive stages. First stage is achieved through the incorporation of local spatial interaction among adjacent pixels in the fuzzy membership function imposed by an auxiliary variable associated with each pixel. The variable describes the involvement level of each pixel for construction of membership functions and different clusters. Then, we adapted a kernel-induced distance to replace the original Euclidean distance in the FCM, which is shown to be more robust than FCM. The problem of sensitivity to noise and intensity inhomogeneity in MRI data is effectively reduced by incorporating a kernel-induced distance metric and local spatial information into a weighted membership function. The experimental results show that the proposed algorithm has advantages in accuracy and robustness against noise in comparison with the FCM, SFCM and CSFCM methods on MRI brain images.

The Causes of Urban Sprawl in Spanish Urban Areas: A Spatial Approach

This paper explores the presence of local spatial interaction on urban development patterns of Spanish urban areas and whether this interaction might be due to the strategic behaviour of neighbouring local governments. Using remote-sensing data from aerial photography and satellite imaging along with Geographic Information Systems (GIS) techniques, it studies urban development patterns across the country with unprecedented detail. The results confirm the main hypothesis that spatial interaction exists in the levels of sprawl between neighbouring municipalities, suggesting that local governments do indeed compete for the creation of new suburban settlement developments, hence promoting excessive urban sprawl.

Difference-in-differences techniques for spatial data: Local autocorrelation and spatial interaction

We consider treatment effect estimation via a difference-in-difference approach for spatial data with local spatial interaction such that the potential outcome of observed units depends on their own treatment as well as on the treatment status of proximate neighbors. We show that under standard assumptions (common trend and ignorability) a straightforward spatially explicit version of the benchmark difference-in-differences regression is capable of identifying both direct and indirect treatment effects. We demonstrate the finite sample performance of our spatial estimator via Monte Carlo simulations.

Difference-in-Differences Techniques for Spatial Data: Local Autocorrelation and Spatial Interaction

We consider treatment effect estimation via a difference-in-difference approach for data with local spatial interaction such that the outcome of observed units depends on their own treatment as well as on the treatment status of proximate neighbors. We show that under standard assumptions (common trend and ignorability) a straightforward spatially explicit version of the benchmark difference-in-differences regression is capable of identifying both direct and indirect treatment effects. We demonstrate the finite sample performance of our spatial estimator via Monte Carlo simulations.

Characteristics of neighborhood interaction in urban land-use changes: A comparative study between three metropolitan areas of Japan

Local spatial interaction between neighborhood land-use categories (i.e. neighborhood interaction) is an important factor which affects urban land-use change patterns. Therefore, it is a key component in cellular automata (CA)-based urban geosimulation models towards the simulation and forecast of urban land-use changes. Purpose of this paper is to interpret the similarities and differences of the characteristics of neighborhood interaction in urban land-use changes of different metropolitan areas in Japan for providing empirical materials to understand the mechanism of urban land-use changes and construct urban geosimulation models. Characteristics of neighborhood interaction in urban land-use changes of three metropolitan areas in Japan, i.e. Tokyo, Osaka, and Nagoya, were compared using such aids as the neighborhood interaction model and similarity measure function. As a result, urban land-use in the three metropolitan areas was found to have had similar structure and patterns during the study period. Characteristics of neighborhood interaction in urban land-use changes are quite different from land-use categories, meaning that the mechanism of urban land-use changes comparatively differs among land-use categories. Characteristics of neighborhood interaction reveal the effect of spatial autocorrelation in the spatial process of urban land-use changes in the three metropolitan areas, which correspond with the characteristics of agglomeration of urban land-use allocation in Japan. Neighborhood interaction amidst urban land-use changes between the three metropolitan areas generally showed similar characteristics. The regressed neighborhood interaction coefficients in the models may represent the general characteristics of neighborhood effect on urban land-use changes in the cities of Japan. The results provide very significant materials for exploring the mechanism of urban land-use changes and the construction of universal urban geosimulation models which may be applied to any city in Japan.

Research of malware propagation in complex networks based on 1-D cellular automata

In this paper, based on 1-D cellular automata, the probabilistic behaviors of malware propagation in complex networks are investigated. Neighborhood and state transition functions with integrated expression are established and two models of malware propagation are proposed to evaluate the probabilistic behavior of malware propagation in various networks. We run the proposed models on nearest-neighbor coupled network （NC） and Erdos-Renyi （ER） random graph network and Watts-Strogatz（WS） small world network and Barabasi-Albert （BA） power law network respectively. Analysis and simulations show that, the proposed models describe perfectly the dynamic behaviors of propagation in the above networks. Furthermore, the proposed models describe not only the average tendency of malware propagation but also the rare events such as saturation and extinction of malware, and overcome the limitation occurring in a deterministic model based on mean-field method that describes only the average tendency of malware propagation and neglects the probabilistic event. Meanwhile, the result of simulations shows that the heterogeneity of degree distribution and local spatial interaction are key factors affecting the malware propagation and immunization.

Gaia as a complex adaptive system.

We define the Gaia system of life and its environment on Earth, review the status of the Gaia theory, introduce potentially relevant concepts from complexity theory, then try to apply them to Gaia. We consider whether Gaia is a complex adaptive system (CAS) in terms of its behaviour and suggest that the system is self-organizing but does not reside in a critical state. Gaia has supported abundant life for most of the last 3.8 Gyr. Large perturbations have occasionally suppressed life but the system has always recovered without losing the capacity for large-scale free energy capture and recycling of essential elements. To illustrate how complexity theory can help us understand the emergence of planetary-scale order, we present a simple cellular automata (CA) model of the imaginary planet Daisyworld. This exhibits emergent self-regulation as a consequence of feedback coupling between life and its environment. Local spatial interaction, which was absent from the original model, can destabilize the system by generating bifurcation regimes. Variation and natural selection tend to remove this instability. With mutation in the model system, it exhibits self-organizing adaptive behaviour in its response to forcing. We close by suggesting how artificial life ('Alife') techniques may enable more comprehensive feasibility tests of Gaia.

GIS and microsimulation for local labour market analysis

This paper outlines the findings of ongoing research on ways of employing and combining geographical information systems (GIS) and microsimulation methodologies for the evaluation and analysis of local labour market problems and policies. First, it is shown how data sets from different sources are integrated to create an urban labour market GIS. In the context of this information system, ways of mapping thematically the local labour market demand and supply are presented. In addition, travel-to-work flows are mapped and it is demonstrated how this data can be used to build local labour market Spatial Interaction Models. Further, it is shown how GIS can be combined with microsimulation techniques to highlight urban problems and enhance the analysis and evaluation of potential social and employment policies. In particular, it is shown how GIS and microsimulation can be used as tools in order to analyse a region's economy and labour market and to estimate the degree of labour market segmentation and socio-economic dualism within an urban system. Also, it is outlined how what-if spatial policy analysis of local labour markets can be performed (i.e. simulating new policy initiatives, firm closures, changes in benefit policies and performing regional multiplier analysis). Finally, the paper presents outputs from SimLeeds, which is a spatial microsimulation model for the Leeds labour market, and explores the potential of GIS combined with microsimulation modelling to create a new framework for the formulation, analysis and evaluation of socio-economic policies at the individual or household level.

The spatiotemporal range of inhibitory interaction in flicker detection

The flicker thresholds of luminous bars were measured as a function of the spatial and/or temporal separation of two flickering stimuli. Each of the bars had an intensity profile of one-half cycle of a sinusoidal wave subtending 2.26 × 0.45 arc deg and each bar was presented twice at two positions with a duration of 10 msec. The spatial separation was defined as the distance between the adjacent flanks of two flickering stimuli, while the temporal separation was determined as the time-lag between the offset of the first flickering stimulus and the onset of the second. We found that the thresholds increased asymptotically with the spatial separation in such a way as to suggest that the spatial extent over which inhibitory interaction could be effective was as large as about 2 arc deg. We also found that the threshold gradually decreased with greater temporal separation; this indicated that the temporal proximity of successive stimuli effects less suppression on the temporal response. These two effects were seemingly additive. These findings suggest that the visual system involves not only local spatial interaction, but also a global mechanism capable of spreading inhibition over several local units after a delay of several msec.

Conditional probability approach to spatial interaction

The merits of a conditional probability (or Markovian) approach to the analysis of local spatial interaction lie in its intuitive appeal and its ability to cope with non-Gaussian situations. Whilst relationships between plausible spatial-temporal models and Markov fields may sometimes be obtained, this does not remove the need for a purely spatial approach to be available. In addition to promoting the above point of view, some Monte Carlo simulation results, due to Mr. D. H. Green, were presented. Analytical progress on coding efficiency was reported.