Change Detection Performance Research Articles

Visual attention-based siamese CNN with SoftmaxFocal loss for laser-induced damage change detection of optical elements

With high-energy laser irradiating, the laser-induced damages may occur in the surfaces of optical elements in laser facilities. As the laser-induced damage changes can badly affect regular and healthy operation of laser facilities, it is essential to effectively detect real damage changes while suppressing meaningless and spurious changes in captured optical images. In order to achieve high-precision laser-induced damage change detection, this paper presents a novel deep learning model which exploits visual attention-based siamese convolutional neural network with SoftmaxFocal loss and significantly improves the performance of damage change detection. In the proposed model, an end-to-end classification network is designed and trained which fuses the spatial-channel domain collaborative attention modules into siamese convolutional neural network thus achieving more efficient feature extraction and representation. For the purpose of addressing the unbalanced distribution of hard and easy samples, a novel loss function which is termed as SoftmaxFocal loss is put forward to train the proposed network. The SoftmaxFocal loss creatively introduces an additive focusing term into original softmax loss which greatly enhances the online hard sample mining ability of the proposed model. Experiments conducted on three real datasets demonstrate the validity and superiority of the proposed model.

Dynamic Task Scheduling for Human-Robot Collaboration

In human-robot collaboration systems, the human-robot team works together to complete all tasks of the assigned job. Tasks’ precedence order is a factor that determines which task should be completed first. A human agent has an expected performance on each task, but the actual performance of the human agent on a task may differ from the expected performance at different times. Therefore, it is also necessary to monitor the human agent so that the task can be transferred to or from the robot when the actual performance of the human agent is worse or better than the expected performance. We propose an architecture for dynamic task allocation and scheduling in a collaborative system. The method is to first identify the group of tasks that should be executed first based on the precedence order of tasks. After obtaining this group of prioritized tasks, we use the quality metric (a numerical value that indicates which agent is better at completing the task) to determine which agent should be assigned each of the tasks. Finally, real-time human agent monitoring is used to detect performance changes in the human agent so that tasks can be transferred from one agent to another. All of these processes are performed automatically without the need for a human agent to schedule or reschedule. We have validated the effectiveness of the architecture through several experiments.

Superpixel-based multiobjective change detection based on self-adaptive neighborhood-based binary differential evolution

With strong penetrability and high resolution, synthetic aperture radar (SAR) images have been widely used in remote sensing image change detection. With the essence of heuristics, metaheuristics are suitable for the needs of most real-life optimization problems according to the expected solution quality and allowable calculation time. To improve the accuracy of change detection, this paper proposes a novel efficient metaheuristic change detection procedure. More specifically, we develop a superpixel-based multiobjective change detection method based on superpixel-wise feature representation and self-adaptive neighborhood-based binary difference evolution. During the clustering for feature analysis, a multiobjective optimization problem (MOP) is modeled which take the likelihood and Bhattacharyya distance between changed and unchanged classes as two objective functions. To solve the MOP, this paper firstly applies the superpixel binary representation technique into the encoding process of multiobjective evolutionary algorithm so that the dimension of pixel coding space can be substantially reduced. Then a self-adaptive neighborhood-based binary differential evolution strategy is proposed to explore the optimal change detection strategy, where a novel mutation operator integrating neighborhood information is designed for the improvement of change detection performance. Experimental results on three real SAR datasets confirm that the proposed method can make a better performance on change detection and the metaheuristic algorithm has good convergence and stability.

A high-resolution feature difference attention network for the application of building change detection

Deep learning based change detection has brought a significant improvement in the accuracy and efficiency when compared with conventional machine learning methods. However, the issues of the lack of differential information and the diversity of the scale features of artificial objects are crucial barriers to the application of building change detection algorithms. A novel deep learning based approach named the high-resolution feature difference attention network (HDANet) is proposed in this work to solve these issues. HDANet can handle the change characteristics well, due to the Siamese network structure. To tackle the loss of the spatial features of buildings caused by the multiple successive down-sampling operations in the current change detection algorithms using fully convolutional networks (FCNs), a multi-resolution parallel structure is introduced in HDANet, and the image information with different resolutions is comprehensively employed, without any spatial information loss. Moreover, an innovative difference attention module is elaborated for the enhancement of the sensitivity to difference information, to keep the building change information. The experimental results obtained on building change detection datasets confirm that HDANet can improve the differential feature representation for change detection, and the performance of the building change detection is also superior to that of the other advanced change detection methods.

Changed Detection Based on Patch Robust Principal Component Analysis

Change detection on retinal fundus image pairs mainly seeks to compare the important differences between a pair of images obtained at two different time points such as in anatomical structures or lesions. Illumination variation usually challenges the change detection methods in many cases. Robust principal component analysis (RPCA) takes intensity normalization and linear interpolation to greatly reduce the illumination variation between the continuous frames and then decomposes the image matrix to obtain the robust background model. The matrix-RPCA can obtain clear change regions, but when there are local bright spots on the image, the background model is vulnerable to illumination, and the change detection results are inaccurate. In this paper, a patch-based RPCA (P-RPCA) is proposed to detect the change of fundus image pairs, where a pair of fundus images is normalized and linearly interpolated to expand a low-rank image sequence; then, images are divided into many patches to obtain an image-patch matrix, and finally, the change regions are obtained by the low-rank decomposition. The proposed method is validated on a set of large lesion image pairs in clinical data. The area under curve (AUC) and mean average precision (mAP) of the method proposed in this paper are 0.9832 and 0.8641, respectively. For a group of small lesion image pairs with obvious local illumination changes in clinical data, the AUC and mAP obtained by the P-RPCA method are 0.9893 and 0.9401, respectively. The results show that the P-RPCA method is more robust to local illumination changes than the RPCA method, and has stronger performance in change detection than the RPCA method.

Similar mechanisms of temporary bindings for identity and location of objects in healthy ageing: an eye-tracking study with naturalistic scenes

The ability to maintain visual working memory (VWM) associations about the identity and location of objects has at times been found to decrease with age. To date, however, this age-related difficulty was mostly observed in artificial visual contexts (e.g., object arrays), and so it is unclear whether it may manifest in naturalistic contexts, and in which ways. In this eye-tracking study, 26 younger and 24 healthy older adults were asked to detect changes in a critical object situated in a photographic scene (192 in total), about its identity (the object becomes a different object but maintains the same position), location (the object only changes position) or both (the object changes in location and identity). Aging was associated with a lower change detection performance. A change in identity was harder to detect than a location change, and performance was best when both features changed, especially in younger adults. Eye movements displayed minor differences between age groups (e.g., shorter saccades in older adults) but were similarly modulated by the type of change. Latencies to the first fixation were longer and the amplitude of incoming saccades was larger when the critical object changed in location. Once fixated, the target object was inspected for longer when it only changed in identity compared to location. Visually salient objects were fixated earlier, but saliency did not affect any other eye movement measures considered, nor did it interact with the type of change. Our findings suggest that even though aging results in lower performance, it does not selectively disrupt temporary bindings of object identity, location, or their association in VWM, and highlight the importance of using naturalistic contexts to discriminate the cognitive processes that undergo detriment from those that are instead spared by aging.

Automated Detection of Cognitive Performance and Resilience Changes in Former Professional American Football Players Following the Administration of a Hemp Extract

Automated Detection of Cognitive Performance and Resilience Changes in Former Professional American Football Players Following the Administration of a Hemp Extract

Deep learning for change detection in remote sensing: a review

ABSTRACT A large number of publications have incorporated deep learning in the process of remote sensing change detection. In these Deep Learning Change Detection (DLCD) publications, deep learning methods have demonstrated their superiority over conventional change detection methods. However, the theoretical underpinnings of why deep learning improves the performance of change detection remain unresolved. As of today, few in-depth reviews have investigated the mechanisms of DLCD. Without such a review, five critical questions remain unclear. Does DLCD provide improved information representation for change detection? If so, how? How to select an appropriate DLCD method and why? How much does each type of change benefits from DLCD in terms of its performance? What are the major limitations of existing DLCD methods and what are the prospects for DLCD? To address these five questions, we reviewed according to the following strategies. We grouped the DLCD information assemblages into the four unique dimensions of remote sensing: spectral, spatial, temporal, and multi-sensor. For the extraction of information in each dimension, the difference between DLCD and conventional change detection methods was compared. We proposed a taxonomy of existing DLCD methods by dividing them into two distinctive pools: separate and coupled models. Their advantages, limitations, applicability, and performance were thoroughly investigated and explicitly presented. We examined the variations in performance between DLCD and conventional change detection. We depicted two limitations of DLCD, i.e. training sample and hardware and software dilemmas. Based on these analyses, we identified directions for future developments. As a result of our review, we found that DLCD’s advantages over conventional change detection can be attributed to three factors: improved information representation; improved change detection methods; and performance enhancements. DLCD has to surpass the limitations with regard to training samples and computing infrastructure. We envision this review can boost developments of deep learning in change detection applications.

Hyperspectral image change detection based on active convolutional neural network and spatial–spectral affinity graph learning

The high spectral resolution of hyperspectral image (HSI) provides the possibility to capture the subtle changes associated with land-cover dynamic evolution process. Supervised deep leaning approaches have been extensively applied to HSI change detection task. However, their success is attributed to the large amount of annotated training samples. Moreover, the large receptive field in the convolutional layer and the presence of the pooling layer reduces the spatial resolution of the deepest FCN layer which will make the predicted change map tends to lack fine object boundary details. In order to boost the HSI change detection performance in a reduced labor of annotating data and enhance the object boundary details of the change map, in this paper, we propose a novel HSI change detection method that integrates both iterative active learning and affinity graph learning into a unified framework. The proposed method consists of two major branches, a unary HSI change detection network branch which learns the pixel-wise change probability, and a pairwise HSI affinity graph learning branch that learns the pairwise affinity of the hyperspectral difference image to refine the coarse probabilities through a random walk connection. To actively select the most informative unlabeled samples, we also propose an HSI change detection active learning strategy based on the spectral property of HSI difference image and the refined change probabilities. The procedures are conducted iteratively to obtain better change detection results progressively. Experimental results show that the proposed method can extract accurate subtle change information while properly preserving the edges and textures of the HSIs with significantly fewer labeled data.

SemiSANet: A Semi-Supervised High-Resolution Remote Sensing Image Change Detection Model Using Siamese Networks with Graph Attention

Change detection (CD) is one of the important applications of remote sensing and plays an important role in disaster assessment, land use detection, and urban sprawl tracking. High-accuracy fully supervised methods are the main methods for CD tasks at present. However, these methods require a large amount of labeled data consisting of bi-temporal images and their change maps. Moreover, creating change maps takes a lot of labor and time. To address this limitation, a simple semi-supervised change detection method based on consistency regularization and strong augmentation is proposed in this paper. First, we construct a Siamese nested UNet with graph attention mechanism (SANet) and pre-train it with a small amount of labeled data. Then, we feed the unlabeled data into the pre-trained SANet and confidence threshold filter to obtain pseudo-labels with high confidence. At the same time, we produce distorted images by performing strong augmentation on unlabeled data. The model is trained to make the CD results of the distorted images consistent with the corresponding pseudo-label. Extensive experiments are conducted on two high-resolution remote sensing datasets. The results demonstrate that our method can effectively improve the performance of change detection under insufficient labels. Our methods can increase the IoU by more than 25% compared to the state-of-the-art methods.

FUSING MULTIPLE UNTRAINED NETWORKS FOR HYPERSPECTRAL CHANGE DETECTION

Abstract. Change detection in hyperspectral images is challenging due to the presence of a large number of spectral bands. Due to the differences in band composition, a deep model trained on one hyperspectral sensor cannot be reused on another hyperspectral sensor. This challenge can be tackled by using untrained models as feature extractor for change detection in hyperspectral images. However, results produced by such a strategy may show variance if the untrained model is slightly perturbed. Different change detection maps are produced from different versions of the untrained model. We propose a decision fusion based strategy that can combine such different results and produce a final change detection map that retains the change information from all change maps. This approach improves the change detection performance and also improves reliability of the result. Experimental results on two publicly available hyperspectral datasets show the effectiveness of the proposed approach.

MSGATN: A Superpixel-Based Multi-Scale Siamese Graph Attention Network for Change Detection in Remote Sensing Images

With the rapid development of Earth observation technology, how to effectively and efficiently detect changes in multi-temporal images has become an important but challenging problem. Relying on the advantages of high performance and robustness, object-based change detection (CD) has become increasingly popular. By analyzing the similarity of local pixels, object-based CD aggregates similar pixels into one object and takes it as the basic processing unit. However, object-based approaches often have difficulty capturing discriminative features, as irregular objects make processing difficult. To address this problem, in this paper, we propose a novel superpixel-based multi-scale Siamese graph attention network (MSGATN) which can process unstructured data natively and extract valuable features. First, a difference image (DI) is generated by Euclidean distance between bitemporal images. Second, superpixel segmentation is employed based on DI to divide each image into many homogeneous regions. Then, these superpixels are used to model the problem by graph theory to construct a series of nodes with the adjacency between them. Subsequently, the multi-scale neighborhood features of the nodes are extracted through applying a graph convolutional network and concatenated by an attention mechanism. Finally, the binary change map can be obtained by classifying each node by some fully connected layers. The novel features of MSGATN can be summarized as follows: (1) Training in multi-scale constructed graphs improves the recognition over changed land cover of varied sizes and shapes. (2) Spectral and spatial self-attention mechanisms are exploited for a better change detection performance. The experimental results on several real datasets show the effectiveness and superiority of the proposed method. In addition, compared to other recent methods, the proposed can demonstrate very high processing efficiency and greatly reduce the dependence on labeled training samples in a semisupervised training fashion.

Open-pit mine change detection from high resolution remote sensing images using DA-UNet++ and object-based approach

ABSTRACT Automatic extraction of land use and land cover changes in mining areas is crucial for the supervision of mineral resources. This study proposes an automatic open-pit mine (OM) change detection method via a novel deep learning model, DA-UNet++ (Deformable-Attention-UNet++), and an object-based approach from high resolution remote sensing images. First, by embedding a deformable convolution module and attention mechanism module into UNet++ network, and constructing a skip sampling strategy, DA-UNet++ is proposed to combat the high heterogeneity of shape and scale of ground objects in OM environment and improve the change detection accuracy. Second, an object-based post-processing (OBPP) approach is presented for enhancing raw change detection results of DA-UNet++, which includes object-based expectation maximisation optimisation and morphological close operation. In the experiments, Season-varying Change Detection Dataset is first employed to evaluate the effectiveness of the proposed DA-UNet++ by comparing with other benchmark models. Then, self-made Ordos OM Change Detection Dataset including 38 OMs is utilised to verify the efficacy of the proposed method, and comprehensive metric of 85.6% is achieved. Evaluation results prove that the proposed method provides superior change detection performance compared with several benchmark methods, and has practical significance for OM management and monitoring.

Development of a New MEWMA – Wilcoxon Sign Rank Chart for Detection of Change in Mean Parameter

A nonparametric chart has been accepted and implemented for real world problems, especially, when the distribution of the population is unknown and the parameter could not be estimated. The nonparametric chart can overcome those limitations and it is user-friendly. Consequently, the objective of this research is to develop a new modified exponentially weighted moving average (MEWMA) chart based on Sign Rank statistics, namely MEWMA-SR, and to compare the performance of change detection with the EWMA and MEWMA charts and nonparametric EWMA based on the Sign (EWMA-Sign) and the Sign Rank (EWMA-SR), and MEWMA-Sign charts. The efficiency measurement of charts is commonly performed by average run length (ARL) divided into two states; in control ARL (ARL0) and out of control ARL (ARL1). The numerical results were carried out by Monte Carlo simulation with 105 replication and the best performance of the chart is considered by the minimum value of ARL1. The proposed chart outperforms when a subgroup is small and a magnitude of changes is moderate for Laplace, otherwise the EWMA-SR is superior to small changes. When the observations are from lognormal the MEWMA-SR performs better than EMWA-SR and other charts for moderate to large changes for all sizes of subgroups. Furthermore, the proposed chart is applied to real data set as the S&P 500 index and shows the best performance in detecting a change.

Human visual gamma for color stimuli.

Strong gamma-band oscillations in primate early visual cortex can be induced by homogeneous color surfaces (Peter et al., 2019; Shirhatti and Ray, 2018). Compared to other hues, particularly strong gamma oscillations have been reported for red stimuli. However, precortical color processing and the resultant strength of input to V1 have often not been fully controlled for. Therefore, stronger responses to red might be due to differences in V1 input strength. We presented stimuli that had equal luminance and cone contrast levels in a color coordinate system based on responses of the lateral geniculate nucleus, the main input source for area V1. With these stimuli, we recorded magnetoencephalography in 30 human participants. We found gamma oscillations in early visual cortex which, contrary to previous reports, did not differ between red and green stimuli of equal L-M cone contrast. Notably, blue stimuli with contrast exclusively on the S-cone axis induced very weak gamma responses, as well as smaller event-related fields and poorer change-detection performance. The strength of human color gamma responses for stimuli on the L-M axis could be well explained by L-M cone contrast and did not show a clear red bias when L-M cone contrast was properly equalized.

Assessing community perceptions and response to climate change in nyando catchment using geospatial technology

The world is experiencing climate crisis that is threatening lives. Those in low lands are displaced by floods loosing possessions and food crops; while those in hilly areas experience land degradation and vegetation loss leading to poverty. Monitoring such occurrence through time is inevitable as it enables governments to make critical decisions for environmental conservation, enhanced resilience and minimize risks of food insecurity. This study explored the use of geospatial technology to investigate climate risk indicators within Nyando catchment. Geospatial technology provides a platform to monitor occurrences of extreme weather conditions, facilitates the understanding of end to end temporal changes of the environment and provides an environment for modelling future scenarios. We derived indicators of climate change within Nyando catchment over the period 2013 to 2021 and inferred their impact on livelihoods. We collected data on rainfall, vegetation cover and land use from Landsat 8 30m satellite images and in situ data using the Mobile mapper global positioning system (GPS) within Nyando catchment. We overlaid these data, characterized them and performed change detection using Erdas Imagine software. Results showed 9.9% reduction in forest cover, 54.1% sedimentation of water areas, 3.9% loss of agricultural land and 2.5% increase in built up areas. The main land use in Nyando catchment is agriculture followed with settlement. We inferred that loss of forest cover is the reason for sedimentation of the water areas, subjecting lowlands to flooding. Moreover, conversion of forest land to agricultural land and settlement were fund to be the main drivers for climate crisis. Indeed agro-ecological approaches to farming will increase forest area while maximizing on vertical farming for increased productivity to meet the ever growing demand for food. We recommend adoption of geospatial technology in monitoring of the landscape for informed decision making to avert risks of climate change.

Feature Decomposition-Optimization-Reorganization Network for Building Change Detection in Remote Sensing Images

Building change detection plays an imperative role in urban construction and development. Although the deep neural network has achieved tremendous success in remote sensing image building change detection, it is still fraught with the problem of generating broken detection boundaries and separation of dense buildings, which tends to produce saw-tooth boundaries. In this work, we propose a feature decomposition-optimization-reorganization network for building change detection. The main contribution of the proposed network is that it performs change detection by respectively modeling the main body and edge features of buildings, which is based on the characteristics that the similarity between the main body pixels is strong but weak between the edge pixels. Firstly, we employ a siamese ResNet structure to extract dual-temporal multi-scale difference features on the original remote sensing images. Subsequently, a flow field is built to separate the main body and edge features. Thereafter, a feature optimization module is designed to refine the main body and edge features using the main body and edge ground truth. Finally, we reorganize the optimized main body and edge features to obtain the output results. These constitute a complete end-to-end building change detection framework. The publicly available building dataset LEVIR-CD is employed to evaluate the change detection performance of our network. The experimental results show that the proposed method can accurately identify the boundaries of changed buildings, and obtain better results compared with the current state-of-the-art methods based on the U-Net structure or by combining spatial-temporal attention mechanisms.

Assessing the status of changing regimes of water bodies in Gorakhpur District, Uttar Pradesh, India.

Water bodies play a very important role in maintaining and restoring the ecological balance, but they are one of the most threatened habitats in the world. Anthropogenic intervention is changing the regimes of wetlands almost everywhere particularly in the developing countries. Gorakhpur District is dotted with many big and small flood plain-related water bodies like rivers, streams, tanks, dead arms, oxbow lakes, etc. Some of these water bodies are worst affected and are degraded by encroachment for agriculture and other economic and developmental activities. Channel migration, aggravated by human intervention, on alluvial plain is also very frequent, which has direct impact on the nature of water bodies and land use transformations of the region. In this paper, the authors have made an attempt to (a) bring current geographical and historical background of water bodies/wetlands for the district. It aims to assess long-term (1917-2018) and short-term (pre- and post-monsoon) changes in the water bodies of Gorakhpur District; (b) provide changes in the regime of water bodies/wetlands and their conversion to different types of land use/land cover classes due to human intervention and due to annual rainy season, which inundates a large extent of the area every year; (C) assess the channel characteristics and morphometric analysis of main rivers of the region during the last hundred years. Remote sensing and Geographical Information System (GIS) have been used to prepare the inventory and to perform change detection, using land use/land cover maps. The floodplain areas of water bodies have almost changed their morphological characters due to encroachment by the nearby areas. Canals, drainage channels, and lakes are the most affected water bodies in the region, which have recorded - 65.38% and 43.37% loss in their area. Even permanent rivers have recorded a decrease of - 16.96% in the area. As per the seasonal change, agriculture land suffered the greatest conversion (18.33%) due to floodwater inundation. The study provides a platform to planners to chalk out their policies and also for monitoring the water bodies. Furthermore, analysis on channel migration will help predict the future course of the main rivers.

SAR Image Change Detection Based on Heterogeneous Graph With Multiattributes and Multirelationships

The performance of change detection between synthetic aperture radar (SAR) images mainly depends on the selection and utilization of image attributes. Nevertheless, most existing change detection approaches merely take the intensity attribute into consideration, constraining their capacities of detecting changes in complex situations. To solve this problem, this study develops an unsupervised SAR image change detection approach based on heterogeneous graph with multi-attributes and multi-relationships. First, the structural attribute, obtained by removing the complex texture, can be used to reflect the overall features of a SAR image. A heterogeneous graph is then constructed to encode the structural and intensity attributes as vertices, and two types of edges are hence connected to capture the intra-relationships and inter-relationships from these different attributes. With the support of this graph, the hyper-adjacency characteristics are proposed to quantify the multi-relationships and describe the global heterogeneous information. Constructing heterogeneous graphs respectively on bi-temporal SAR images, the changes of both intensity and structural attributes can be measured via comparing the hyper-adjacency characteristics of the bi-temporal graphs, thereby generating a difference image with good separability. Finally, the change map is obtained by cutting off the weakly related edges of heterogeneous graph on the difference image. Experiments on four real SAR datasets prove the effectiveness of the proposed graph-driven approach in improving the accuracy and robustness of change detection.

Novel Multiscale Decision Fusion Approach to Unsupervised Change Detection for High-Resolution Images

High-resolution remote sensing images usually contain multiscale information, which can be used to enhance the change detection (CD) performance. How to make effective use of multiscale information needs intensive study. This letter presents a novel multiscale decision fusion (MDF) method for unsupervised CD based on Dempster–Shafer (DS) theory and modified conditional random field (CRF). The method consists of three main steps: 1) images of three different scales are created automatically by image segmentation, and then three-scale difference images (DIs) are produced by applying change vector analysis to the three-scale images; 2) the membership function of each scale DI is estimated by fuzzy clustering, and the fusion membership, as well as an initial CD map, is obtained by combining the estimated membership using DS theory; and 3) the initial CD map is refined with an improved CRF that incorporates a spatial attraction model. The proposed method can combine the multiscale information in images and the spatial contextual information. The effectiveness of the proposed method was validated by two experiments with high-resolution remote sensing images.