Survey on Segmentation and Classification Techniques of Satellite Images by Deep Learning Algorithm

This survey paper aims to show methods to analyze and classify field satellite images using deep learning and machine learning algorithms.Users of deep learning-based Convolutional Neural Network(CNN)technology to harvest fields from satellite images or generate zones of interest were among the planned application scenarios(ROI).Using machine learning,the satellite image is placed on the input image,segmented,and then tagged.In contem-porary categorization,field size ratio,Local Binary Pattern(LBP)histograms,and color data are taken into account.Field satellite image localization has several practical applications,including pest management,scene analysis,and field tracking.The relationship between satellite images in a specific area,or contextual information,is essential to comprehending the field in its whole.

Fusing Satellite Images Using ABC Optimizing Algorithm

Fusing satellite(remote sensing)images is an interesting topic in processing satellite images.The result image is achieved through fusing information from spectral and panchromatic images for sharpening.In this paper,a new algorithm based on based the Artificial bee colony(ABC)algorithm with peak signalto-noise ratio(PSNR)index optimization is proposed to fusing remote sensing images in this paper.Firstly,Wavelet transform is used to split the input images into components over the high and low frequency domains.Then,two fusing rules are used for obtaining the fused images.The first rule is“the high frequency components are fused by using the average values”.The second rule is“the low frequency components are fused by using the combining rule with parameter”.The parameter for fusing the low frequency components is defined by using ABC algorithm,an algorithm based on PSNR index optimization.The experimental results on different input images show that the proposed algorithm is better than some recent methods.

Disaster Monitoring of Satellite Image Processing Using Progressive Image Classification

The analysis of remote sensing image areas is needed for climate detec-tion and management,especially for monitoringflood disasters in critical environ-ments and applications.Satellites are mostly used to detect disasters on Earth,and they have advantages in capturing Earth images.Using the control technique,Earth images can be used to obtain detailed terrain information.Since the acquisi-tion of satellite and aerial imagery,this system has been able to detectfloods,and with increasing convenience,flood detection has become more desirable in the last few years.In this paper,a Big Data Set-based Progressive Image Classification Algorithm(PICA)system is introduced to implement an image processing tech-nique,detect disasters,and determine results with the help of the PICA,which allows disaster analysis to be extracted more effectively.The PICA is essential to overcoming strong shadows,for proper access to disaster characteristics to false positives by operators,and to false predictions that affect the impact of the disas-ter.The PICA creates tailoring and adjustments obtained from satellite images before training and post-disaster aerial image data patches.Two types of proposed PICA systems detect disasters faster and more accurately(95.6%).

Integrating Highly Spatial Satellite Image for 3D Buildings Modelling Using Geospatial Algorithms and Architecture Environment

The growing demand for current and precise geographic information that pertains to urban areas has given rise to a significant interest in digital surface models that exhibit a high level of detail. Traditional methods for creating digital surface models are insufficient to reflect the details of earth’s features. These models only represent three-dimensional objects in a single texture and fail to offer a realistic depiction of the real world. Furthermore, the need for current and precise geographic information regarding urban areas has been increasing significantly. This study proposes a new technique to address this problem, which involves integrating remote sensing, Geographic Information Systems (GIS), and Architecture Environment software environments to generate a detailed three-dimensional model. The processing of this study starts with: 1) Downloading high-resolution satellite imagery; 2) Collecting ground truth datasets from fieldwork; 3) Imaging nose removing; 4) Generating a Two-dimensional Model to create a digital surface model in GIS using the extracted building outlines; 5) Converting the model into multi-patch layers to construct a 3D model for each object separately. The results show that the 3D model obtained through this method is highly detailed and effective for various applications, including environmental studies, urban development, expansion planning, and shape understanding tasks.

Comparison of Machine Learning Methods for Satellite Image Classification: A Case Study of Casablanca Using Landsat Imagery and Google Earth Engine

Satellite image classification is crucial in various applications such as urban planning,environmental monitoring,and land use analysis.In this study,the authors present a comparative analysis of different supervised and unsupervised learning methods for satellite image classification,focusing on a case study in Casablanca using Landsat 8 imagery.This research aims to identify the most effective machine-learning approach for accurately classifying land cover in an urban environment.The methodology used consists of the pre-processing of Landsat imagery data from Casablanca city,the authors extract relevant features and partition them into training and test sets,and then use random forest(RF),SVM(support vector machine),classification,and regression tree(CART),gradient tree boost(GTB),decision tree(DT),and minimum distance(MD)algorithms.Through a series of experiments,the authors evaluate the performance of each machine learning method in terms of accuracy,and Kappa coefficient.This work shows that random forest is the best-performing algorithm,with an accuracy of 95.42%and 0.94 Kappa coefficient.The authors discuss the factors of their performance,including data characteristics,accurate selection,and model influencing.

Study of the propagation direction of the internal waves in the South China Sea using satellite images

Internal wave propagation carries considerable vertical shear which can lead to turbulence and mixing. Based on the analysis of more than 2 500 synthetic aperture radar （SAR） and optical satellite images, the in- ternal wave propagation in the whole South China Sea was investigated systematically. The results show that （1） in the northeastern South China Sea, most internal waves propagate westward from the Luzon Strait and are diffracted by coral reefs near the Dongsha Islands. Some impinge onto the shelf and a few are reflected; （2） in the northwestern South China Sea, most internal waves are generated at the shelf and propagate northwestward or westward to the coast; （3） in the western South China Sea, most internal waves propagate westward to the Vietnamese coast, except a few propagate southward to the deep sea; and （4） in the southern South China Sea, most internal waves propagate southwestward to the coast. Some prop- agate southeastward to the coast of Kalimantan Island, and a few propagate southeastward because of the influence of the Mekon~ River.

Level study on fractal characteristics of tidal creeks and information of seashell habitats in the Gaizhou Beach based on high-resolution satellite images

The fractal characteristics of tidal creeks in the Gaizhou Beach are analyzed based on high-resolution images fusionof Landsat TM and ERS2, and then the graphic models and characteristics of converse information tree of tidalcreeks in the Gaizhou Beach are established. A calculation model is established based on the above results, and at thesame time, quantitative calculation of the evolution characteristics and the diversity between the northern and thesouthern parts of the Gaizhou Beach is carried out. By the supervised classification of these images, distribution andareas of high tidal flats, middle tidal flats and low tidal flats in the Gaizhou Beach are studied quantitatively, and imagecharactistics of seashell habitats in the Gaizhou Beach and the correlation between mudflat distribution and seashellhabitats are studied. At last, the engineering problems in the Gaizhou Beach are discussed.

Quantifying post logging biomass loss using satellite images and ground measurements in Southeast Cameroon

Forest logging in the Congo Basin has led to forest fragmentation due to logging infrastructures and felling gaps. In the same vein, forest concessions in the Congo Basin have increasing interest in the REDD＋ mechanism. However, there is little information or field data on carbon emissions from forest degradation caused by logging. To help fill this gap, Landsat 7 and 8 and SPOT 4 images of the East Region of Cameroon were processed and combined with field measurements （measurement of forest roads widths, felling gaps and log yards） to assess all disturbed areas. Also, measurements of different types of forest infrastructures helped to highlight emission factors. Forest contributes to 5.18 % of the degradation of the annual allowable cut （AAC） （84.53 ha） corresponding to 4.09 % of forest carbon stock （6.92 t ha^-1）. Felling gaps constitute the primary source of degradation, represented an estimated area of 32.41 ha （2 % of the cutting area） far ahead of primary roads （18.44ha） and skid trails （16.36 ha）. Assessment of the impact of degradation under the canopy requires the use of high resolution satellite images and field surveys.

Comparison of satellite images with different spatial resolutions to estimate stand structural diversity in urban forests

The structural diversity in urban forests is highly important to protect biodiversity. In particular, fruit trees and bush species, cavity-bearing trees and coarse, woody debris provide habitats for animals to feed, nest and hide. Improper silvicultural practices, intensive recreational use and illegal harvesting lead to a decline in the structural diversity in forests within larger metropolitan cities. It is important to monitor the structural diversity at definite time intervals using effective technologies with a view to instituting the necessary conservation measures. The use of satellite images seems to be appropriate to this end. Here we aimed to identify the associations between the textural features derived from the satellite images with different spatial resolutions and the structural diversity indices in urban forest stands (Shannon-Wiener index, complexity index, dominance index and density of wildlife trees). RapidEye images with a spatial resolution of 5?m × 5?m, ASTER images with a spatial resolution of 15?m × 15?m and Landsat-8 ETM satellite images with a spatial resolution of 30?m × 30?m were used in this study. The first-order (standard deviation of gray levels) and second order (GLCM entropy, GLCM contrast and GLCM correlation) textural features were calculated from the satellite images. When associations between textural features in the images and the structural diversity indices were assessed using the Pearson correlation coefficient, very high associations were found between the image textural features and the diversity indices. The highest association was found between the standard deviation of gray levels (SDGLRAP) derived from RVIRAP of RapidEye image and the Shannon-Wiener index (H h) calculated on the basis of tree height (R 2?=?0.64). The findings revealed that RapidEye satellite images with a spatial resolution of 5?m × 5?m are most suitable for estimating the structural diversity in urban forests.

Integrating cross-sensor high spatial resolution satellite images to detect subtle forest vegetation change in the Purple Mountains,a national scenic spot in Nanjing,China

Accurate information on the location and magnitude of vegetation change in scenic areas can guide the configuration of tourism facilities and the formulation of vegetation protection measures.High spatial resolution remote sensing images can be used to detect subtle vegetation changes.The major objective of this study was to map and quantify forest vegetation changes in a national scenic location,the Purple Mountains of Nanjing,China,using multi-temporal cross-sensor high spatial resolution satellite images to identify the main drivers of the vegetation changes and provide a reference for sustainable management.We used Quickbird images acquired in 2004,IKONOS images acquired in 2009,and WorldView2 images acquired in 2015.Four pixel-based direct change detection methods including the normalized difference vegetation index difference method,multi-index integrated change analysis(MIICA),principal component analysis,and spectral gradient difference analysis were compared in terms of their change detection performances.Subsequently,the best pixel-based detection method in conjunction with object-oriented image analysis was used to extract subtle forest vegetation changes.An accuracy assessment using the stratified random sampling points was conducted to evaluate the performance of the change detection results.The results showed that the MIICA method was the best pixel-based change detection method.And the object-oriented MIICA with an overall accuracy of 0.907 and a kappa coefficient of 0.846 was superior to the pixel-based MIICA.From 2004 to 2009,areas of vegetation gain mainly occurred around the periphery of the study area,while areas of vegetation loss were observed in the interior and along the boundary of the study area due to construction activities,which contributed to 79%of the total area of vegetation loss.During 2009–2015,the greening initiatives around the construction areas increased the forest vegetation coverage,accounting for 84%of the total area of vegetation gain.In spite of this,vegetation loss occurred in the interior of the Purple Mountains due to infrastructure development that caused conversion from vegetation to impervious areas.We recommend that:(1)a local multi-agency team inspect and assess law enforcement regarding natural resource utilization;and(2)strengthen environmental awareness education.

Retrieval of High Resolution Satellite Images Using Texture Features

In this research, a content-based image retrieval （CBIR） system for high resolution satellite images has been developed by using texture features. The proposed approach uses the local binary pattern （LBP） texture feature and a block based scheme. The query and database images are divided into equally sized blocks, from which LBP histograms are extracted. The block histograms are then compared by using the Chi-square distance. Experimental results show that the LBP representation provides a powerful tool for high resolution satellite images （HRSI） retrieval.

Assessment of theoretical approaches to derivation of internal solitary wave parameters from multi-satellite images near the Dongsha Atoll of the South China Sea

This study assesses the accuracy and the applicability of the Korteweg-de Vries(KdV)and the nonlinear Schr?dinger(NLS)equation solutions to derivation of dynamic parameters of internal solitary waves(ISWs)from satellite images.Visible band images taken by five satellite sensors with spatial resolutions from 5 m to 250 m near the Dongsha Atoll of the northern South China Sea(NSCS)are used as a baseline.From the baseline,the amplitudes of ISWs occurring from July 10 to 13,2017 are estimated by the two approaches and compared with concurrent mooring observations for assessments.Using the ratio of the dimensionless dispersive parameter to the square of dimensionless nonlinear parameter as a criterion,the best appliable ranges of the two approaches are clearly separated.The statistics of total 18 cases indicate that in each 50%of cases,the KdV and the NLS approaches give more accurate estimates of ISW amplitudes.It is found that the relative errors of ISW amplitudes derived from two theoretical approaches are closely associated with the logarithmic bottom slopes.This may be attributed to the nonlinear growth of ISW amplitudes as propagating along a shoaling thermocline or topography.The test results using three consecutive satellite images to retrieve the ISW propagation speeds indicate that the use of multiple satellite images(>2)may improve the accuracy of retrieved phase speeds.Meanwhile,repeated multi-satellite images of ISWs can help to determine the types of ISWs if mooring data are available nearby.

Derivation of sea surface current field from sequential satellite images of the East China Sea

A series of NOAA AVHRR data over the East China Sea were collected from the ground station of the Second Institute of Oceanography, Hangzhou, China. Three methods, including a functional analytic method (FAM), a maximum cross correlation (MCC)'method and a correlation relaxation (C - R) method, are applied to derive the sea surface current field from sequential satellite images in the area of the East China Sea. Several preprocessing steps, such as geometric correction, SST determination, image projection, image navigation and grey value normalization as well as land and cloud mask are performed. The results from the three methods reflect the general current system in this area reasonably.

The digital mapping of satellite images under no ground control and the distribution of landform, blue ice and meteorites in the Grove Mountains, Antarctica

The colorful satellite image maps with the scale of 1∶100000 were made by processing the parameters-on-satellite under the condition of no data of field surveying. The purpose is to ensure the smooth performance of the choice of expedition route, navigation and research task before the Chinese National Antarctic Research Expedition (CHINARE) first made researches on the Grove Mountains. Moreover, on the basis of the visual interpretation of the satellite image, we preliminarily analyze and discuss the relief and landform, blue ice and meteorite distribution characteristics in the Grove Mountains.

Analysis of Some Characteristics of the Fault Activities in Eastern China by Satellite Images

Active faults have special electromagnetic effect and remote sensing characteristics, and exhibit unique im-agery marks in satellite images. A comprehensive comparison of images of active faults in eastern China and ananalysis of geologic and geomorphic data can tell us some characteristics of fault activity in the area during theneotectonic period: 1) The fault activities of the north-south tectonic zone, North China and Taiwan werestronger than those of southeastern and northeastern China; 2) the faulting in the north-south tectonic zone,North China and Taiwan has continued up to now, and most of the fault activites in southeastern andnorth-eastern China have become weaker since the Middle Pleistocene; 3) the activity is unsteady in time, mostbeing intermittent, or episodic, i.e. alternately strong and weak; 4) most active faults of a definite size can be di-vided into several segments which somewhat differ from each other in the characteristics of the activity.

Using U-Net to Detect Buildings in Satellite Images

This report presented a method that uses deep computing and stochastic gradient descent algorithm to automatically detect building from satellite images. In this method, a convolutional neural network architecture called U-Net was trained to highlight the building pixels from the rest of the image. This method applied a binary cross-entropy loss function, used ADAM algorithm for gradient descent optimization, and adopted interaction-over-union for accuracy measurement. Continuous loss decreases and accuracy increases were observed during the training and validation. Finally, the visualization of the predicted masks from the trained model after 20 epochs proved that the U-Net model delivers over 60% Intersection over Union accuracy results for detecting buildings from satellite images.

Structural Interpretation of Lineaments by Satellite Image Processing(Landsat TM)in the Region of Zahret Medien(Northern Tunisia)

There are several techniques that were developed for determining the linear features. Lineament extraction?from satellite data has been the most widely used applications in geology. In the present study, lineament has?been extracted from the digital satellite scene (Landsat 5, TM data), in the region of Zahret Median situated in the north west of Tunisia. The image was enhanced and used for automatic extraction. Several directions of features were mapped. The directions of major invoices are NE-SW and NW-SE oriented. The validation of the obtained results is carried out by comparison with the results geophysics as well as to the studies previous of mapping developed in the sector of study.

Estimating Rice Yield by HJ-1A Satellite Images

As illustrated by the case of Xuyi County, Jinhu County and Hongze County in Jiangsu Province, China, monitoring and forecasting of rice production were carried out by using HJ-1A satellite remote sensing images. The handhold GPS machines were used to measure the geographical position and some other information of these samples such as area shape. The GPS data and the interpretation marks were used to correct H J-1 image, assist human-computer interactive interpretation, and other operations. The test data had been participated in the whole classification process. The accuracy of interpreted information on rice planting area was more than 90% By using the leaf area index from the normalized difference vegetation index inversion, the biomass from the ratio vegetation index inversion, and combined with the rice yield estimation model, the rice yield was estimated. Further, the thematic map of rice production classification was made based on the rice yield data. According to the comparison results between measured and fitted values of yields and areas of sampling sites, the accuracy of the yield estimation was more than 85%. The results suggest that HJ-A/B images could basically meet the demand of rice growth monitoring and yield forecasting, and could be widely applied to rice production monitoring.

An Adaptive and Image-guided Fusion for Stereo Satellite Image Derived Digital Surface Models

The accuracy of Digital Surface Models(DSMs)generated using stereo matching methods varies due to the varying acquisition conditions and configuration parameters of stereo images.It has been a good practice to fuse these DSMs generated from various stereo pairs to achieve enhanced,in which multiple DSMs are combined through computational approaches into a single,more accurate,and complete DSM.However,accurately characterizing detailed objects and their boundaries still present a challenge since most boundary-ware fusion methods still struggle to achieve sharpened depth discontinuities due to the averaging effects of different DSMs.Therefore,we propose a simple and efficient adaptive image-guided DSM fusion method that applies k-means clustering on small patches of the orthophoto to guide the pixel-level fusion adapted to the most consistent and relevant elevation points.The experiment results show that our proposed method has outperformed comparing methods in accuracy and the ability to preserve sharpened depth edges.

U-Net Inspired Deep Neural Network-Based Smoke Plume Detection in Satellite Images

Industrial activities, through the human-induced release of Green House Gas (GHG) emissions, have beenidentified as the primary cause of global warming. Accurate and quantitative monitoring of these emissions isessential for a comprehensive understanding of their impact on the Earth’s climate and for effectively enforcingemission regulations at a large scale. This work examines the feasibility of detecting and quantifying industrialsmoke plumes using freely accessible geo-satellite imagery. The existing systemhas so many lagging factors such aslimitations in accuracy, robustness, and efficiency and these factors hinder the effectiveness in supporting timelyresponse to industrial fires. In this work, the utilization of grayscale images is done instead of traditional colorimages for smoke plume detection. The dataset was trained through a ResNet-50 model for classification and aU-Net model for segmentation. The dataset consists of images gathered by European Space Agency’s Sentinel-2 satellite constellation from a selection of industrial sites. The acquired images predominantly capture scenesof industrial locations, some of which exhibit active smoke plume emissions. The performance of the abovementionedtechniques and models is represented by their accuracy and IOU (Intersection-over-Union) metric.The images are first trained on the basic RGB images where their respective classification using the ResNet-50model results in an accuracy of 94.4% and segmentation using the U-Net Model with an IOU metric of 0.5 andaccuracy of 94% which leads to the detection of exact patches where the smoke plume has occurred. This work hastrained the classification model on grayscale images achieving a good increase in accuracy of 96.4%.