Multispectral Classification Research Articles

Use of Remote Sensing and Geographic Information System for the Analysis of Urban Development: A Case Study of Banyumas Regency, Indonesia

The change in land-use/land-cover (LULC) is one of the physical factors in urban development and the rapid growth in population has necessitates the need for space, driven by high socio-economic activities. Banyumas is a regency that had experienced rapid population growth in the last two decades, establishing an activity and service center in the Barlingmascakeb region. This rapid population growth has led to massive changes in LULC. Therefore, this study aimed to observe the changes in LULC in 2000 and 2020 to determine the direction of urban development in Banyumas Regency within 20 years. Multispectral classification with a Maximum Likelihood algorithm was used to extract LULC information from Landsat images. The changes obtained by crosstab analysis on the multispectral classification results were used as a reference to observe the direction of urban development. This procedure used four quadrants according to the cardinal directions and Standard Deviational Ellipse (SDE). The result showed that LULC in the forest class experienced the highest change of 142,584.3 km², accounting for 48%. Based on the increase in built-up land over 20 years, the direction of urban development according to the cardinal directions showed that the most dominant increase was in quadrant II (Southeast), which is 56.44% or 21.95 km². It was concluded that the direction of urban development was oriented toward the southeast.

A Semantically Nonredundant Continuous-Scale Feature Network for Panchromatic and Multispectral Classification

A Semantically Nonredundant Continuous-Scale Feature Network for Panchromatic and Multispectral Classification

Dual Branch Feature Representation and Variational Autoencoder for Panchromatic and Multispectral Classification

Dual Branch Feature Representation and Variational Autoencoder for Panchromatic and Multispectral Classification

Multi-Sensor and Multi-Scale Remote Sensing Approach for Assessing Slope Instability along Transportation Corridors Using Satellites and Uncrewed Aircraft Systems

Rapid slope instabilities (i.e., rockfalls) involving highway networks in mountainous areas pose a threat to facilities, settlements and life, thus representing a challenge for asset management plans. To identify different morphological expressions of degradation processes that lead to rock mass destabilization, we combined satellite and uncrewed aircraft system (UAS)-based products over two study sites along the State Highway 133 sector near Paonia Reservoir, Colorado (USA). Along with a PS-InSAR analysis covering the 2017–2021 interval, a high-resolution dataset composed of optical, thermal and multi-spectral imagery was systematically acquired during two UAS surveys in September 2021 and June 2022. After a pre-processing step including georeferencing and orthorectification, the final products were processed through object-based multispectral classification and change detection analysis for highlighting moisture or lithological variations and for identifying areas more susceptible to deterioration and detachments at the small and micro-scale. The PS-InSAR analysis, on the other hand, provided multi-temporal information at the catchment scale and assisted in understanding the large-scale morpho-evolution of the displacements. This synergic combination offered a multiscale perspective of the superimposed imprints of denudation and mass-wasting processes occurring on the study site, leading to the detection of evidence and/or early precursors of rock collapses, and effectively supporting asset management maintenance practices.

Land-Cover Change Detection in Batur Catchment Area Using Remote Sensing

Land cover information is an essential aspect in the planning and management of earth modeling and understanding. Land cover changes impact the physical and social environment, such as hydrological conditions and ecological systems. This study aimed to identify spatial differences in the land cover of the Batur catchment area from 2015-2021 by using a remote sensing approach to describe the existing land-cover site and to detect its changes. The methods used in this study are a combination of the vegetation index and a supervised classification maximum likelihood algorithm with Landsat 8 OLI/TIRS in 2015 and 2021. Furthermore, the Change Detection Feature, identified from two image periods in 2015-2021 and processed, is used to detect changes in land cover. The accuracy assessment utilized QuickBird imagery recorded in 2015; field survey data were taken in 2021. The results showed that between 2015 to 2021, built-up area, bare land, shrubs, and lake have increased by 102,66% (306,01 ha), 27,95% (452,25 ha), 15,20% (215,72 ha) and 4,05 % (62,73 ha) while dryland forest and dry-dry-field have decreased by -25,84% (-606,29 ha) and -14.59% (-430,42 ha), respectively. The overall accuracy of the multispectral classification results in 2015 and 2021 was 82,63% and 89,57%.Keywords: Land-Cover Change; Batur; Catchment Area; Remote SensingÂ

ESTIMATION OF LAND SURFACE TEMPERATURE-ASSESSMENT WITH REMOTE SENSING DATA FOR URBAN HEAT ISLAND IN BATAM MUNICIPALITY

Urban Heat Island (UHI) is a phenomenon where the surface temperature in urban areas is higher than in the surrounding area or in open-air areas. Hilly city continues to increase every year. Because of that, it needed an effort at the city-level urban heat island on Batam Island. Through the Thermal Landsat 8 channel image recording May 24, 2018 sensing data can be used to complete the heat calculation of the urban heat Island on Batam Island. This research aims to determine the variation of Urban Heat Island (UHI) effect on land cover in the research area. This research uses LST and maximum likelihood multispectral classification for the land cover area and NDVI transformation. The results of the processing are obtained with an R2 value of 0.872324 with a UHI intensity of 0-8.06 ̊ C. This shows the amount needed by Urban Heat Cities in Batam Island specifically in urban areas such as Batu Ampar District, Batam Kota, Bengkong, Batu Aji, Lubuk Baja, Sagulung, and Sekupang

Mapping build-up area density using normalized difference built-up index (ndbi) and urban index (ui) wetland in the city banjarmasin

Density and undeveloped land can be identified using digital transformation through remote sensing data. This study aims to map the distribution of building densities with the Normalized Difference Built-Up Index (NDBI) and Urban Index (UI), analyze the comparison of building densities of the NDBI and UI methods, and analyze the relationship between the transformation of NDBI and UI on building densities in Banjarmasin City. The data used to obtain building density image Landsat 8 Oli Tirs. The method used to separate the built area and non-building area using digital classification. The area was developed from the results of the multispectral classification filtered with NDBI and UI transformation for the classification of the building density level. The results of this study indicate that building density has a positive relationship with the transformation of NDBI and UI because the high building density in the transformation of NDBI and UI has a high value. Non-built area and constructed land shows an NDBI accuracy of 91.4505% and 92.5359% accuracy of UI, there is a difference of 1.0854% greater accuracy UI. Overall accuracy exceeding 80% indicates very high accuracy for building density mapping.

Biomass Calculations of Individual Trees Based on Unmanned Aerial Vehicle Multispectral Imagery and Laser Scanning Combined with Terrestrial Laser Scanning in Complex Stands

Biomass is important in monitoring global carbon storage and the carbon cycle, which quickly and accurately estimates forest biomass. Precision forestry and forest modeling place high requirements on obtaining the individual parameters of various tree species in complex stands, and studies have included both the overall stand and individual trees. Most of the existing literature focuses on calculating the individual tree species’ biomass in a single stand, and there is little research on calculating the individual tree biomass in complex stands. This paper calculates the individual tree biomass of various tree species in complex stands by combining multispectral and light detection and ranging (LIDAR) data. The main research steps are as follows. First, tree species are classified through multispectral data combined with field investigations. Second, multispectral classification data are combined with LIDAR point cloud data to classify point cloud tree species. Finally, the divided point cloud tree species are used to compare the diameter at breast height (DBH) and height of each tree species to calculate the individual tree biomass and classify the overall stand and individual measurements. The results show that under suitable conditions, it is feasible to identify tree species through multispectral classification and calculate the individual tree biomass of each species in conjunction with point-cloud data. The overall accuracy of identifying tree species in multispectral classification is 52%. Comparing the DBH of the classified tree species after terrestrial laser scanning (TLS) and unmanned aerial vehicle laser scanning (UAV-LS) to give UAV-LS+TLS, the concordance correlation coefficient (CCC) is 0.87 and the root-mean-square error (RMSE) is 10.45. The CCC and RMSE are 0.92 and 1.41 compared with the tree height after UAV-LS and UAV-LS+TLS.

Distribution analysis of coral reefs for development of marine tourism in Weh Island, Aceh, Indonesia

One of the most attractive ecosystems to be used as tourist destinations is coral reefs. The coral reef ecosystem on Weh Island, Aceh, Indonesia is one of the coastal ecosystems that have a strategic role in ecological and economic development. One of the ecological and economic development efforts can be done through marine tourism. This study aims to analyze the distribution of coral reefs to develop marine tourism in Weh Island. This study uses Landsat 8 OLI image data and field observations. The technique used in this research is the image data analysis technique using multispectral classification. The results showed that the coral reefs on Weh Island in 2020 amounted to 13,136,000 Ha. Therefore, the development of marine tourism must create tourism zones to maintain the sustainability of coral reef ecosystems.

A unified heuristic approach to simultaneously detect fusarium and ergot damage in wheat

Fusarium spp. and Ergot (Claviceps purpurea) are fungal pathogenic species that infect cereals worldwide, causing widespread economic losses. Fusarium-damaged kernels (FDKs) in wheat often contain harmful mycotoxins, while Ergot damage (ERG) is commonly characterized by the presence of alkaloids. Both FDK and ERG damage could cause health hazards to people and animals. Currently, assessment of FDKs and ERG is performed by human visual inspection using distinctly different biochemical assays and/or wet chemistry methods, all of which are subjective, time-consuming and non-repeatable. In this study, the authors present a unified non-destructive optical method using visible-near-infrared (VNIR) (400-1000 nm) and short-wavelength near-infrared (SWIR) (1000-2500 nm) hyperspectral imaging (HI) to simultaneously detect FDK and ERG damage in wheat kernels. The un-mixing technique of spectral angle mapping (SAM) was employed for pixel- and object-wise classification of the three classes i.e. FDK, ERG and control (healthy), of non-touching wheat kernels. Overall classification accuracies of VNIR–93% and SWIR–85% were obtained with 100% classification of ERG samples in the VNIR range. To reduce model complexity, principle component analysis (PCA) was explored, and a subset of 12 significant wavelengths was selected. The SAM model based on these selected wavelengths did not achieve accurate prediction, indicating it requires the entire spectral data for reliable performance. However, a partial least square discriminant analysis (PLSDA) model using the reduced wavelength subset achieved overall classification accuracies of VNIR–94% and SWIR–90%, which were better when compared to the SAM method. This study is the first known attempt to simultaneously identify two important fungal diseases in wheat using a completely non-invasive and non-destructive method. The results confirm the feasibility of developing a fast and cost-effective multi-spectral classification technique for FDK and ERG detection.

Classification of airborne multispectral imagery to quantify common vole impacts on an agricultural field.

BACKGROUNDThe common vole (Microtus arvalis) is a very destructive agricultural pest. Particularly in Europe, its monitoring is essential not only for adequate management and outbreak forecasting, but also for accurately determining the vole's impact on affected fields. In this study, several alternatives for estimating the damage to alfalfa fields by voles through unmanned vehicle systems (UASs) and multispectral cameras are presented. Currently, both the farmers and agencies involved in the integrated pest management (IPM) programs of voles do not have sufficiently precise methods for accurate assessments of the real impact to crops.RESULTSOverall, the four multispectral classification methods presented showed similar performances. However, the normalized difference vegetation index (NDVI)‐based segmentation exhibited the most accurate and reliable appraisal of the affected areas. Nevertheless, it must be noted that the simplest method, which was based on an automatic classification, provided results similar to those obtained by more complex methods. In addition, a significant direct relationship was found between the number of active burrows and damage to the alfalfa canopy.CONCLUSIONUnmanned vehicle systems, combined with multispectral imagery classification, are an effective and easily transferable methodology for the assessment and monitoring of common vole damage to agricultural plots. This combination of methods facilitates decision‐making processes for IPM control strategies against this pest. © 2022 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Analysis on the change of runoff curve number influence to surface flow debit using ALOS AVNIR-2 data imagery

One part of the hydrologic cycle which has a major influence in increasing the amount of river flow discharge is surface runoff. The higher surface runoff discharge, causing the possibility of surface flooding, therefore required an empirical model that can calculate the amount of surface runoff so as to produce updated data and quickly change according to their needs. One of the empirical methods that can be used to calculate the amount of surface runoff is by using the curve number method. This research is done by utilizing remote sensing image, that is, ALOS AVNIR-2. Data extraction from ALOS imagery includes land cover information using multispectral classification analysis, slope inclination information through visual interpretation, and land use interpretation. The runoff that occurred in Banjarnegara Regency tends to be high, that is, 61.24 percent of the total area of the research area. Large runoff with very high/extreme class spread on the form of hilly land to the old volcano complex at the study site. The runoff in the medium to low class only covers 3.56 percent of the total area and is distributed on the fluvial form with the flat-to-flat slopes. The result of analysis of runoff data is obtained from slope analysis and type of land use in the research location. Increasingly steep slope with little vegetation-land use, then the greater the runoff that occurs. Finally, the research result could be implemented into higher student class activity, especially in remote sensing classes, GIS, cloud computing, and big data analysis. By this process, the students will be improved their skills in analyzing imagery data as well as create new information derived from the remote sensing data.

Adaptive Dual-Path Collaborative Learning for PAN and MS Classification

Due to the limitation of sensor technology, researchers tend to obtain high-quality image information from panchromatic (PAN) images and multispectral (MS) images with different resolutions. Therefore, the classification of remote sensing images of PAN and MS have become a research hotspot. In this paper, we propose an adaptive dual-path collaborative learning method for PAN and MS classification. In the stage of sample generation and training, we propose an adaptive neighborhood sample grading (ANSG) strategy in the establishing sample stage so that each pixel to be classified can obtain neighborhood information suitable for itself. Further, to simulate biological cognitive mechanisms, we divide the samples into different levels, and design the self-paced progressive loss (SPL), thus allowing the network to do preference training in different stages. The network’s training can quickly reach the optimal of the current stage and the overall convergence is more thorough. In the network structure, we propose a dual-path module (DPM) to effectively alleviate the gradient degradation in the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">residual path</i> , while ensuring maximum gradient loss information flow between every two layers in the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">densely connected path</i> . This module can extract more robust features to cope with the complex characteristics of remote-sensing images. Moreover, using the characteristics of the dual path to better fuse the features by the gradual collaborative fusion (GCF) way. The experimental results and theoretical analysis have demonstrated the proposed approach’s effectiveness, feasibility, and robustness. Our model are available at https://github.com/AIpy-nan/DBFI-Net.

A Simple Data Preprocessing and Postprocessing Techniques for SVM Classifier of Remote Sensing Multispectral Image Classification

Present scenario of the remote sensing domain deals with how to utilize the data for different purposes like classification, target detection, disaster management, change detection, flood monitoring, deforestation, etc. Now due to improvements in the sensor technology very high spatial and spectral resolutions data are available. Over a decade various new advanced research papers have been projected in the literature for spatial and spectral classification of such high-resolution remote sensing images. Thematic information investigation of the earth&#x0027;s surface image is possible by the classification technique and the most frequently used method for this purpose is multispectral classification using a supervised learning process. In the supervised learning process, the specialist challenges to discover exact sites in the remotely sensed data that represent homogeneous examples of the known land cover type. The most recommended method for the classification of remote sensing (RS) images is the support vector machine (SVM) because of its high accuracy but any classifier depends on good quality training samples. The collection of authentic training samples of different classes is a critical issue when the whole classification result is important. This paper presents a pre-processing technique based on local statistics for generation-correction of training samples with quadrant division. A simple filter-based post-processing technique is proposed for the improvement of classification accuracy. We study rigorously how the proposed pre-processing technique has affected the result of classification accuracy for different kernels SVM classifiers. Also, we have presented the comparison results between the proposed method and other different classifiers in the literature.

Novel Cross-Resolution Feature-Level Fusion for Joint Classification of Multispectral and Panchromatic Remote Sensing Images

With the increasing availability and resolution of satellite sensor data, multispectral (MS) and panchromatic (PAN) images are the most popular data that are used in remote sensing among applications. This article proposes a novel cross-resolution hidden layer feature fusion (CRHFF) approach for joint classification of multiresolution MS and PAN images. In particular, shallow spectral and spatial features at a global scale are first extracted from an MS image. Then, deep cross-resolution hidden layer features extracted from MS and PAN are fused from patches at a local scale according to an autoencoder (AE)-like deep network. Finally, the selected multiresolution hidden layer features are classified in a supervised manner. By taking advantage of integrated shallow-to-deep and global-to-local features from the high-resolution MS and PAN images, the cross-resolution latent information can be extracted and fused in order to better model imaged objects from the multimodal representation and finally increase the classification accuracy. Experimental results obtained on three real multiresolution datasets covering complex urban scenarios confirm the effectiveness of the proposed approach in terms of higher accuracy and robustness with respect to literature methods.

A Two-Stage Mutual Fusion Network for Multispectral and Panchromatic Image Classification

With the rapid development of remote sensing technology, satellites can easily obtain multispectral (MS) and panchromatic (PAN) images. How to mine the essence and peculiarity of the MS and PAN images and utilize their complementary to improve classification performance is still a challenge. This paper designs a two-stage mutual fusion network (TSMF-Net) for MS and PAN image classification. The network can be divided into two stages: data fusion and feature fusion. In the data fusion stage, we propose an adaptive twin intensity-hue-saturation (ATIHS) strategy. It not only aligns the size and channels of the MS and PAN images by a novel q-Split operation, but also introduces an adaptive soft-average mask to reduce the differences between replacement components, effectively mitigating spectral distortion and paving the way for the next stage. In the feature fusion stage, we propose a feature graft block (FG-Block) in which we introduce triplet loss and design an interlaced channel addition (ICA) module. Under the supervision of triplet loss, the FG-Block separates and hauls each branch’s essential and peculiar features. With the help of the ICA module, it can effectively graft the essential feature between branches and retain the peculiar feature of each branch, improving the utilization and discrimination of features. Finally, composed of the ATIHS, FG-Blocks, and output layers, our TSMF-Net is proven to improve the accuracy of the remote sensing classification task. The experimental results on multiple datasets verify the effectiveness of our proposed algorithms. Our code is available at: https://github.com/liaoyinuo/TSMF-Net.

GCFnet: Global Collaborative Fusion Network for Multispectral and Panchromatic Image Classification

Among various multimodal remote sensing data, the pairing of multispectral (MS) and panchromatic (PAN) images is widely used in remote sensing applications. This article proposes a novel global collaborative fusion network (GCFnet) for joint classification of MS and PAN images. In particular, a global patch-free classification scheme based on an encoder-decoder deep learning (DL) network is developed to exploit context dependencies in the image. The proposed GCFnet is designed based on a novel collaborative fusion architecture, which mainly contains three parts: 1) two shallow-to-deep feature fusion branches related to individual MS and PAN images; 2) a multiscale cross-modal feature fusion branch of the two images, where an adaptive loss weighted fusion strategy is designed to calculate the total loss of two individual and the cross-modal branches; 3) a probability weighted decision fusion strategy for the fusion of the classification results of three branches to further improve the classification performance. Experimental results obtained on three real datasets covering complex urban scenarios confirm the effectiveness of the proposed GCFnet in terms of higher accuracy and robustness compared to existing methods. By utilizing both sampled and non-sampled position data in the feature extraction process, the proposed GCFnet can achieve excellent performance even in a small sample-size case. The codes will be available from the website: https://github.com/SicongLiuRS/GCFnet.

Binary Neural Network for Multispectral Image Classification

Compared with traditional images, multispectral images (MSIs) contain more spectral bands and higher data dimensions. The existing MSI classification model has high computational complexity and consumes a lot of computing resources. In this letter, we propose a lightweight multispectral classification method named CABNN based on binary neural networks (BNNs) to effectively have a trade-off between model performance and computational cost. First, we modify and binarize the MobileNetV1 network and add almost computation-free shortcuts to enhance the expressive capability. Secondly, since the BNN is sensitive to the distribution of activation functions, we introduce RPReLU with learnable coefficients to automatically adjust activation distribution at almost no extra cost. Lastly, considering that MSIs have multiple channels, we utilize an efficient channel attention (ECA) module to assign different weights to each channel to concentrate on crucial features and suppress insignificant features. We conduct experiments on four public MSI datasets, including NaSC-TG2, EuroSAT, GID Fine land-cover classification, and UC Merced Land Use. Extensive experiments demonstrate that the proposed CABNN has higher efficiency and better comprehensive performance than the state-of-the-art methods across the board.

Hyperspectral and Multispectral Classification for Coastal Wetland Using Depthwise Feature Interaction Network

The monitoring of coastal wetlands is of great importance to the protection of marine and terrestrial ecosystems. However, due to the complex environment, severe vegetation mixture, and difficulty of access, it is impossible to accurately classify coastal wetlands and identify their species with traditional classifiers. Despite the integration of multisource remote sensing data for performance enhancement, there are still challenges with acquiring and exploiting the complementary merits from multisource data. In this article, the depthwise feature interaction network (DFINet) is proposed for wetland classification. A depthwise cross attention module is designed to extract self-correlation and cross correlation from multisource feature pairs. In this way, meaningful complementary information is emphasized for classification. DFINet is optimized by coordinating consistency loss, discrimination loss, and classification loss. Accordingly, DFINet reaches the standard solution-space under the regularity of loss functions, while the spatial consistency and feature discrimination are preserved. Comprehensive experimental results on two hyperspectral and multispectral wetland datasets demonstrate that the proposed DFINet outperforms other competitive methods in terms of overall accuracy.

Pemetaan Distribusi Mangrove Menggunakan Citra Sentinel-2A: Studi Kasus Kota Langsa

Mangroves are one of the most productive ecosystems for human life, marine ecosystems, and coastal areas. Mangrove distribution is a distribution based on specific geographical or administrative boundaries. Kota Langsa is one of the areas that has a good representation of the distribution of mangroves. Therefore, researchers studied the Kota Langsa area because Kota Langsa is one of the areas with the largest and most diverse mangrove ecosystem in Aceh Province. This study examines the mapping of mangrove distribution using Sentinel-2A multispectral imagery with composite images of Red, Green, and Blue. This research uses SNAP software. The research stages consist of radiometric correction, atmospheric correction, and multispectral image classification. The method used in image classification is the maximum likelihood algorithm. The use of the maximum likelihood algorithm is because the maximum likelihood algorithm gives the best results among other algorithms. The development of the research is the distribution of mangroves in Langsa City, covering an area of 4727.35 ha, which is divided into three sub-districts and eleven gampong (kelurahan). The sub-districts that have mangrove distribution are East Langsa District covering an area of 3240.25 Ha (68.55%), Langsa Barat District covering an area of 1486.47 Ha (31.45%), and Langsa Lama District covering an area of 0.63 Ha (0.013).