scholarly journals Examining the Deep Belief Network for Subpixel Unmixing with Medium Spatial Resolution Multispectral Imagery in Urban Environments

2019 ◽  
Vol 11 (13) ◽  
pp. 1566 ◽  
Author(s):  
Yingbin Deng ◽  
Renrong Chen ◽  
Changshan Wu
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Spatial Resolution ◽  
Landsat Imagery ◽  
Absolute Error ◽  
Urban Environments ◽  
Deep Belief Network ◽  
Support Vector ◽  
Belief Network ◽  
Mixed Pixel

Mixed pixels in medium spatial resolution imagery create major challenges in acquiring accurate pixel-based land use and land cover information. Deep belief network (DBN), which can provide joint probabilities in land use and land cover classification, may serve as an alternative tool to address this mixed pixel issue. Since DBN performs well in pixel-based classification and object-based identification, examining its performance in subpixel unmixing with medium spatial resolution multispectral image in urban environments would be of value. In this study, (1) we examined DBN’s ability in subpixel unmixing with Landsat imagery, (2) explored the best-fit parameter setting for the DBN model and (3) evaluated its performance by comparing DBN with random forest (RF), support vector machine (SVM) and multiple endmember spectral mixture analysis (MESMA). The results illustrated that (1) DBN performs well in subpixel unmixing with a mean absolute error (MAE) of 0.06 and a root mean square error (RMSE) of 0.0077. (2) A larger sample size (e.g., greater than 3000) can provide stable and high accuracy while two-RBM-layer and 50 batch sizes are the best parameters for DBN in this study. Epoch size and learning rate should be decided by specific applications since there is not a consistent pattern in our experiments. Finally, (3) DBN can provide comparable results compared to RF, SVM and MESMA. We concluded that DBN can be viewed as an alternative method for subpixel unmixing with Landsat imagery and this study provides references for other scholars to use DBN in subpixel unmixing in urban environments.

scholarly journals Fusion Methods and Multi-classifiers to Improve Land Cover Estimation Using Remote Sensing Analysis

2021 ◽  
Author(s):  
Hayder Dibs ◽  
Hashim Ali Hasab ◽  
Ammar Shaker Mahmoud ◽  
Nadhir Al-Ansari
Keyword(s):  
Remote Sensing ◽  
Land Use ◽  
Land Cover ◽  
Spatial Resolution ◽  
Kappa Coefficient ◽  
Support Vector ◽  
Land Use Land Cover ◽  
Multispectral Imagery ◽  
Spectral Angle Mapper ◽  
Spectral Angle

AbstractAdopting a low spatial resolution remote sensing imagery to get an accurate estimation of Land Use Land Cover is a difficult task to perform. Image fusion plays a big role to map the Land Use Land Cover. Therefore, This study aims to find out a refining method for the Land Use Land Cover estimating using these steps; (1) applying a three pan-sharpening fusion approaches to combine panchromatic imagery that has high spatial resolution with multispectral imagery that has low spatial resolution, (2) employing five pixel-based classifier approaches on multispectral imagery and fused images; artificial neural net, support vector machine, parallelepiped, Mahalanobis distance and spectral angle mapper, (3) make a statistical comparison between image classification results. The Landsat-8 image was adopted for this research. There are twenty Land Use Land Cover thematic maps were generated in this study. A suitable and reliable Land Use Land Cover method was presented based on the most accurate results. The results validation was performed by adopting a confusion matrix method. A comparison made between the images classification results of multispectral imagery and all fused images levels. It proved the Land Use Land Cover map produced by Gram–Schmidt Pan-sharpening and classified by support vector machine method has the most accurate result among all other multispectral imagery and fused images that classified by the other classifiers, it has an overall accuracy about (99.85%) and a kappa coefficient of about (0.98). However, the spectral angle mapper algorithm has the lowest accuracy compared to all other adopted methods, with overall accuracy of 53.41% and the kappa coefficient of about 0.48. The proposed procedure is useful in the industry and academic side for estimating purposes. In addition, it is also a good tool for analysts and researchers, who could interest to extend the technique to employ different datasets and regions.

scholarly journals Kesesuaian dan Arahan Penggunaan Lahan Berdasarkan Rencana Pola Ruang Wilayah di Hulu Daerah Aliran Sungai Kelara

2017 ◽  
Vol 9 (2) ◽  
pp. 75
Author(s):  
Usman Arsyad ◽  
Andang Suryana Soma ◽  
Wahyuni Wahyuni ◽  
Tita Rahayu Arief
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Spatial Pattern ◽  
Rice Field ◽  
Landsat Imagery ◽  
Agroforestry Systems ◽  
Community Forest ◽  
Dryland Agriculture ◽  
Dry Land ◽  
Land Cover Maps

This study aimed to analyze the compatibility between the land cover spatial pattern plan and determine the direction of land use in the event of a discrepancy. This research was conducted on the Kelara Upstream Watershed located in gowa and jeneponto using land cover maps generated from landsat imagery interpretation 8. Then overlay to map the spatial pattern plan. Then determined the order of land use is done when there is a discrepancy between the results of the overlay with maps of land cover spatial pattern plan. The result showed that 41,05% of the total area of the Kelara Upstream Watershed of 28.185,68 ha a land use form of a orchards. After overlay discovered discrepancy land cover maps with maps of spatial pattern plan. Based on a map spatial pattern plan that should in reality the field is man made forest, orchards, dryland agriculture and rice field. According to these condition the specified order of land use that is Hkm (Community Forest) with agroforestry and Agroforestry Systems. Rice field In the Protected and Production forest order to intensification land use and plantations forest, orchards and dry land agriculture order to Community Forest with agroforestry systems . In the area of cultivation the land use rice field, orchards and dryland agriculture order to agroforestry systems.

scholarly journals A Land Cover Changes Model Based on Carbon Stock in the Mount Patah Region of, Bengkulu-Indonesia

2020 ◽  
Vol 4 (9) ◽  
pp. 1-5
Keyword(s):  
Land Use ◽  
Information System ◽  
Land Cover ◽  
Forest Inventory ◽  
Carbon Stock ◽  
Forest Cover ◽  
Landsat Imagery ◽  
Land Cover Changes ◽  
Land Area ◽  
Reduced Data

Forest cover in Bengkulu is reduced. Data from WARSI shows, 1990 forest cover areas in the province are approximately 1,009,209 hectares or 50.4 % of the land area reaching 1,979,515 hectares. But now, it is only 685,762 hectares of the area of his blood. That is, the period of 25 years, there is a forest cover decline of 323,447 hectares. Forest and land cover changes are the largest contributor to greenhouse gas emissions. The purpose of this article is to see land cover changes based on carbon stock in the years 2009 and 2018. Model of land cover change based on carbon stock year 2028 and 2038. The method of this research uses the calculation of the Stock Difference Approach with spatial analysis of national land closure of Landsat imagery 2009-2018 and biomass data for forest inventory results Geographic Information System (GIS). The results of this research were the reduced forest area and the change in land use changed from 2009 and 2018. So carbon stock is also increasingly reduced.

scholarly journals IMPACT ANALYSIS OF URBANIZATION ON LAND USE LAND COVER CHANGE FOR KHULNA CITY, BANGLADESH USING TEMPORAL LANDSAT IMAGERY

2018 ◽  
Vol XLII-5 ◽  
pp. 757-760 ◽  
Author(s):  
M. Moniruzzam ◽  
A. Roy ◽  
C. M. Bhatt ◽  
A. Gupta ◽  
N. T. T. An ◽  
...  
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Impact Analysis ◽  
Landsat Imagery ◽  
Training Dataset ◽  
Anthropogenic Factors ◽  
Vacant Land ◽  
Landsat Data ◽  
Land Use Land Cover ◽  
Decadal Scale

<p><strong>Abstract.</strong> Urbanization has given a massive pace in Land Use Land Cover (LULC) changes in rapidly growing cities like Khulna, i.e. the third largest city of Bangladesh. Such impacting changes have taken place in over-decadal scale. It is important because detailed analysis with regularly monitoring will be fruitful to drag the attention of decision maker and urban planner for sustainable development and to overcome the problem of urban sprawl. In this present study, changes in LULC as an impact of urbanization, have been investigated for years 1997, 2002, 2007, 2012 and 2017; using three generation of Landsat data in geographic information system (GIS) domain which has the height competence in recent time. Initially, LULC have categorised into Built-up, Vegetation, Vacant Land, and Waterbody with the help of supervised classification technique. Field work had been carried out for acquiring training dataset and validation. The accuracy has been achieved more than 85% for the changes assessed. Analysis has an outlet with increase in built-up area by 27.92% in year 1997 to 2017 and continued respectively in each successive interval of half a decade at the given years. On the other side waterbody and vacant land decreased correspondingly. Bound to mention, instead to having largest temporal durability, the moderate spatial resolution of Landsat data has a limitation for such urban studies. These changes are responsible by both of natural or anthropogenic factors. Such study will provide a better way out of optimization of land-use to prepare detail area plan (DAP) of Khulna City Corporation (KCC) and Khulna development authority (KDA).</p>

scholarly journals Analysis of land use/land cover change in Adei watershed, Central Highlands of Ethiopia

2019 ◽  
Vol 41 (1) ◽  
pp. 146-153 ◽  
Author(s):  
Megersa Olumana Dinka ◽  
Degefa Dhuga Chaka
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Supervised Classification ◽  
Landsat Imagery ◽  
Land Cover Changes ◽  
Land Use Land Cover ◽  
Post Processing ◽  
Ancillary Data ◽  
The Stability ◽  
Land Cover Maps

Abstract Land use/land cover changes (LULCC) at Adei watershed (Ethiopia) over a period of 23 years (1986–2009) has been analysed from LANDSAT imagery and ancillary data. The patterns (magnitude and direction) of LULCC were quantified and the final land use/land cover maps were produced after a supervised classification with appropriate post-processing. Image analysis results revealed that the study area has undergone substantial LULCC, primarily a shift from natural cover into managed agro-systems, which is apparently attributed to the increasing both human and livestock pressure. Over the 23 years, the aerial coverage of forest and grass lands declined by 8.5% and 4.3%, respectively. On the other hand, agricultural and shrub lands expanded by 9.1% and 3.7%, respectively. This shows that most of the previously covered by forest and grass lands are mostly shifted to the rapidly expanding farm land use classes. The findings of this study suggested that the rate of LULCC over the study period, particularly deforestation due to the expansion of farmland need to be given due attention to maintain the stability and sustainability of the ecosystem.

scholarly journals Impacts of Aerosol Particle Size Distribution and Land Cover Land Use on Precipitation in a Coastal Urban Environment Using a Cloud-Resolving Mesoscale Model

2014 ◽  
Vol 2014 ◽  
pp. 1-17 ◽  
Author(s):  
Nathan Hosannah ◽  
Jorge E. Gonzalez
Keyword(s):  
Land Use ◽  
Particle Size ◽  
Land Cover ◽  
Aerosol Particle ◽  
Mesoscale Model ◽  
Cloud Condensation Nuclei ◽  
Atmospheric Modeling ◽  
Urban Environments ◽  
Precipitation Event ◽  
Aerosol Particle Size

Urban environments influence precipitation formation via response to dynamic effects, while aerosols are intrinsically necessary for rainfall formation; however, the partial contributions of each on urban coastal precipitation are not yet known. Here, the authors use aerosol particle size distributions derived from the NASA aerosol robotic network (AERONET) to estimate submicron cloud condensation nuclei (CCN) and supermicron CCN (GCCN) for ingestion in the regional atmospheric modeling system (RAMS). High resolution land data from the National Land Cover Database (NLCD) were assimilated into RAMS to provide modern land cover and land use (LCLU). The first two of eight total simulations were month long runs for July 2007, one with constant PSD values and the second with AERONET PSDs updated at times consistent with observations. The third and fourth runs mirrored the first two simulations for “No City” LCLU. Four more runs addressed a one-day precipitation event under City and No City LCLU, and two different PSD conditions. Results suggest that LCLU provides the dominant forcing for urban precipitation, affecting precipitation rates, rainfall amounts, and spatial precipitation patterns. PSD then acts to modify cloud physics. Also, precipitation forecasting was significantly improved under observed PSD and current LCLU conditions.

scholarly journals Fast outlier detection for high-dimensional data of wireless sensor networks

2020 ◽  
Vol 16 (10) ◽  
pp. 155014772096383
Author(s):  
Yan Qiao ◽  
Xinhong Cui ◽  
Peng Jin ◽  
Wu Zhang
Keyword(s):  
Wireless Sensor Networks ◽  
Support Vector Machine ◽  
Sensor Networks ◽  
Outlier Detection ◽  
Large Scale ◽  
Deep Belief Network ◽  
Wireless Sensor ◽  
High Dimensional ◽  
Support Vector ◽  
Belief Network

This article addresses the problem of outlier detection for wireless sensor networks. As increasing amounts of observational data are tending to be high-dimensional and large scale, it is becoming increasingly difficult for existing techniques to perform outlier detection accurately and efficiently. Although dimensionality reduction tools (such as deep belief network) have been utilized to compress the high-dimensional data to support outlier detection, these methods may not achieve the desired performance due to the special distribution of the compressed data. Furthermore, because most existed classification methods must solve a quadratic optimization problem in their training stage, they cannot perform well in large-scale datasets. In this article, we developed a new form of classification model called “deep belief network online quarter-sphere support vector machine,” which combines deep belief network with online quarter-sphere one-class support vector machine. Based on this model, we first propose a model training method that learns the radius of the quarter sphere by a sorting method. Then, an online testing method is proposed to perform online outlier detection without supervision. Finally, we compare the proposed method with the state of the arts using extensive experiments. The experimental results show that our method not only reduces the computational cost by three orders of magnitude but also improves the detection accuracy by 3%–5%.

scholarly journals Landscape-wide changes in land use and land cover correlate with, but rarely explain local biodiversity change

2020 ◽  
Vol 35 (10) ◽  
pp. 2255-2273
Author(s):  
Martin Jung ◽  
Jörn P. W. Scharlemann ◽  
Pedram Rowhani
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Functional Groups ◽  
Relative Abundance ◽  
Photosynthetic Activity ◽  
Landsat Imagery ◽  
Geometric Mean ◽  
Bird Diversity ◽  
Assemblage Composition ◽  
Biodiversity Change

Abstract Context There is an ongoing debate whether local biodiversity is declining and what might drive this change. Changes in land use and land cover (LULC) are suspected to impact local biodiversity. However, there is little evidence for LULC changes beyond the local scale to affect biodiversity across multiple functional groups of species, thus limiting our understanding of the causes of biodiversity change. Objectives Here we investigate whether landscape-wide changes in LULC, defined as either trends in or abrupt changes in magnitude of photosynthetic activity, are driving bird diversity change. Methods Linking 34 year (1984–2017) time series at 2745 breeding bird survey (BBS) routes across the conterminous United States of America with remotely-sensed Landsat imagery, we assessed for each year what proportion of the landscape surrounding each BBS route changed in photosynthetic activity and tested whether such concomitant or preceding landscape-wide changes explained changes in bird diversity, quantified as relative abundance (geometric mean) and assemblage composition (Bray–Curtis index). Results We found that changes in relative abundance was negatively, and assemblage composition positively, correlated with changes in photosynthetic activity within the wider landscape. Furthermore, landscape-wide changes in LULC in preceding years explained on average more variation in bird diversity change than concomitant change. Overall, landscape-wide changes in LULC failed to explain most of the variation in bird diversity change for most BBS routes regardless whether differentiated by functional groups or ecoregions. Conclusions Our analyses highlight the influence of preceding and concomitant landscape-wide changes in LULC on biodiversity.

scholarly journals Online Recognition Method for Voltage Sags Based on a Deep Belief Network

Energies ◽  
2018 ◽  
Vol 12 (1) ◽  
pp. 43 ◽  
Author(s):  
Fei Mei ◽  
Yong Ren ◽  
Qingliang Wu ◽  
Chenyu Zhang ◽  
Yi Pan ◽  
...  
Keyword(s):  
Large Scale ◽  
Deep Belief Network ◽  
Support Vector ◽  
Voltage Sag ◽  
Voltage Sags ◽  
Data Sampling ◽  
Recognition Method ◽  
Belief Network ◽  
Online Recognition ◽  
Residential Electricity

Voltage sag is a serious power quality phenomenon that threatens industrial manufacturing and residential electricity. A large-scale monitoring system has been established and continually improved to detect and record voltage sag events. However, the inefficient process of data sampling cannot provide valuable information early enough for governance of the system. Therefore, a novel online recognition method for voltage sags is proposed. The main contributions of this paper include: 1) The causes and waveform characters of voltage sags were analyzed; 2) according to the characters of different sag waveforms, 10 voltage sag characteristic parameters were proposed and proven to be effective; 3) a deep belief network (DBN) model was built using these parameters to complete automatic recognition of the sag event types. Experiments were conducted using voltage sag data from one month recorded by the 10 kV monitoring points in Suqian, Jiangsu Province, China. The results showed good performance of the proposed method: Recognition accuracy was 96.92%. The test results from the proposed method were compared to the results from support vector machine (SVM) recognition methods. The proposed method was shown to outperform SVM.

Sign in / Sign up

Export Citation Format

Share Document