Sara Khanbani; Reza Shahhoseini
Abstract
Extended AbstractIntroductionChange detection (CD) from remote sensing image considered an important topic among scientist because of its application in monitoring urban and non-urban area, environmental issue, damage assessment, etc. Presenting an efficient CD method from a high-resolution image can ...
Read More
Extended AbstractIntroductionChange detection (CD) from remote sensing image considered an important topic among scientist because of its application in monitoring urban and non-urban area, environmental issue, damage assessment, etc. Presenting an efficient CD method from a high-resolution image can face a different challenge; most of the CD method from a high-resolution image requires training procedure to overcome this challenge. In this paper, an unsupervised (without needing training process) CD algorithm proposed from the high-resolution image. In this method spatial and spectral features extracted from bi-temporal images of the studied area. Difference images generated from high information content features. Then generated different images mapped into spherical space. The Primary change map created using implemented multi-thresholding method on created spherical space and the second change map created using hierarchical clustering regularized by Markov random field method. The final change map created by integrating the result of primary and secondary change maps. The final change map shows an overall accuracy of 92.56% in the studied area. Data and methodsThe data used in this paper is a subset of the main data with dimensions of 2000 * 2000 from an urban area in the city of Mashhad. These images corresponded to the two periods of 1390 and 1395 and were taken with UAV. The orthoimage is related to the first time with a spatial resolution of 6 cm and the second image is taken with a pixel size of 10 cm. In this paper, in order to detect of change of high-resolution images, first, the input images are registered in terms of spectral and spatial, and then feature images are extracted from each input image separately. In the next step, the differences images corresponding to high information content feature images are calculated. . The optimal difference images are mapped to the spherical space using selected statistical methods and in order to better analysis of the results. Otsu multi-thresholding method implemented on r component of sphere space. In the next step, the optimal difference image mapped to a spherical space is divided into non-overlapping blocks with the same dimensions; a cumulative hierarchical clustering method is applied for each block separately. In this case, the computational volume and space proposed in the hierarchical clustering method are reduced. The results of the cumulative clustering of the blocks are merged together and then the Markov random field method is used in order to regularize the results of the cluster in order to reduce noise.In final clustering, the class values below the lowest Otsu threshold are known as unchanged pixels with high reliability and the values above the maximum threshold are determined as changed pixels. The class of middle interval is unknown. For determining, the class of middle interval the corresponded output of hierarchy clustering regularized with a random Markov field is used. In the last step, a vegetation and shadow mask is used for final post-processing. Results and discussionIn order to an accurate assessment of the proposed method on the mentioned study area, a ground truth image with 11073 pixels has been used as a ground test image. The proposed method has shown an overall accuracy of 92.56 in the study area. The accuracy of detecting changed pixels shows 81.61% and the accuracy of detection unchanged pixels shows 92.77%. The false alarm percentage is 0.21 percent and the missed alarm accuracy is 0.0723 percent. For comparative evaluation, the proposed method is compared with the change vector analysis algorithm. In this section, the selected features in the feature extraction section are entered in the change analysis algorithm, and then the multi thresholding algorithm and shadow analysis used to create the final change map. This method has shown increasing the alarm in comparison with the proposed method. The accuracy of changed and un-changed pixels in the change vector analysis method is equal to 52.98 and 89.24%, respectively. Comparing these results with the results of the proposed method shows the efficiency of the proposed method. ConclusionIn this paper, the new unsupervised change detection method presented based on the combination of multi thresholding and the hierarchical clustering algorithm. Compared to supervised methods that require training data, this method does not require training data. In this method, textural and spatial-spectral features are extracted from images with high spatial resolution, which covers the discussion of the importance of neighborhoods in images with high spatial resolution. In the next step, the extracted features that have a high information content are selected, which helps to reduce the redundancy of the information. The contrast images of the features with high information content are created to differentiate the location of the changes. Spherical computing space is considered as the basic computing space. In order to create a binary change map, two analyzes have been performed on the spherical computational space. First, the Otsu multi-thresholding method has been applied. The values of the smaller and larger thresholds have definite classes. But the value of the middle interval needs to be further analyzed using the hierarchical clustering method. In this section, the middle pixel class is examined, and then a final adjustment is performed using Markov field and shadow and vegetation analysis in order to post-process and prevent false changes. In this paper, the parameters of changed accuracy – unchanged accuracy - overall accuracy - false and missed alarms have been used to evaluate the accuracy of the proposed method with a ground accuracy map. In order to make a comparative study, the proposed method is compared with the change vector analysis method of the created feature space. The results show the efficiency of the proposed method.
Hadi Farhadi; Tayebe Managhebi; Hamid Ebadi
Abstract
Extended Abstract1- IntroductionRemote Sensing (RS), as one of the most efficient mapping technologies, is employed in wide areas due to its speed, cost-effectiveness, monitoring over wide areas and using time series data. So far, several data and methods are used for this purpose. In general, RS active ...
Read More
Extended Abstract1- IntroductionRemote Sensing (RS), as one of the most efficient mapping technologies, is employed in wide areas due to its speed, cost-effectiveness, monitoring over wide areas and using time series data. So far, several data and methods are used for this purpose. In general, RS active and passive sensors provide useful information in various applications such as building extraction, natural resource management, agricultural monitoring, etc. The extraction of accurate information about the location, density and distribution of buildings in the urban areas is one of the major challenges in the urban study which is used in various applications. In this framework, the monitoring of the urban parameters, such as urban green space, public health, and environmental justice, urban density and so on has been accomplished by radar and optical image processing, in the last three decades. So far, various methods, including Artificial Intelligence (AI), Deep Learning (DL), object-based methods, etc. have been proposed to extract information in the urban areas. However, an important issue is access to the powerful computer hardware to process the time-series images. In such a situation, the use of the Google Earth Engine (GEE) as a web-based RS platform and its ability to perform spatial and temporal aggregations on a set of satellite images has been considered by many researchers. In this research, a semi-automatic method was developed building extraction in Tabriz, northwest of Iran, based on the satellite images using the GEE cloud computing platform. Since accessible data is one of the most important challenges in the use of space RS, in this study, the free Sentinel-1 and sentinel-2 data, which belongs to the European Space Agency (ESA), has been utilized. 2- Materials & Methods2-1- Study AreaThe study area is central part of the city of Tabriz East Azerbaijan Province, which is located in northwestern of Iran. 2-2- DataVarious data sources have been used in this study, including Sentinel-1, Sentinel-2, and the Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM). In addition, 400 training samples were created using High-Resolution Google Earth Imagery (GEI) in two classes: urban-residential (buildings) and non-residential areas (vegetation, soil, road, water and etc.).2-3- MethodologyThe goal of this research is to develop a method for identifying the buildings in an urban area. For this purpose, after importing images and pre-processing them in the GEE Platform, a map of the Primary Urban Areas (PUA) and High-Potential Buildings (HPB) was produced from Sentinel-1 images according to the sensitivity of the radar images to the target physical parameters. Then, in order to remove the annoying features and extract the Secondary Urban Areas (SUA), spectral indices such as Normalized Difference Vegetation Index (NDVI), Green Normalized Difference Vegetation Index (GNDVI), Renormalized Difference Vegetation Index (RDVI), Normalized Difference Water Index (NDWI), Soil Extraction Index (SOEI), Normalized Difference Built-up Index (NDBI), and Build-up Extraction Index (BUEI) were extracted from Sentinel-2 images. Also, the high slope of the area and the mountainous areas was extracted from the SRTM DEM data and used as a mask in the final results. Afterwards, the unimodal histogram thresholding method was used in order to determine the threshold value for each index. Finally, by merging the map of HPB and the map of SUA, the final map was produced and evaluated by other methods. In this research, the proposed method used images from GEI with a very high spatial resolution to validate the generated map. As a result, sampling was carried out using a visual interpretation of GEI in two classes: residential areas (buildings) and non-residential areas. The samples were selected randomly and 400 points were collected for each residential and non-residential class. In the study area, a total of 800 test points were used to evaluate the results of the proposed method. To evaluate the accuracy of the results, the criteria of overall accuracy (OA), kappa coefficient (KC), user accuracy (UA) and producer accuracy (PA) were used. 3- Results & DiscussionAccording to the visual interpretation, all buildings in urban areas with a length and width greater than 10 meters (spatial resolution of the four major bands of Sentinel2) can be extracted using the proposed method in this study, and the results are acceptable in various features. According to the proposed method, annoying features such as vegetation and water body areas were removed from the building identification process with high accuracy, and the accuracy in the study area was improved. The results showed that the OA and KC were 90.11 % and 0.803, respectively. Based on the quantitative and qualitative comparisons, the proposed method had a very satisfying performance. 4- ConclusionDue to the spectral diversity and the presence of various features in urban environments, preparing a map related to it in a large area is extremely difficult. In this regard, the current study presented a very fast semi-automatic method for preparing the urban area map and extracting buildings in Tabriz using Sentinel-1 and Sentinel-2 satellite images as a time series in the GEE platform. One of the most significant benefits of the proposed method is that the data and processing system used in our study is free. Thus, in addition to not having to download large amounts of data, the method presented in the current study has the ability to eliminate many of the limitations of traditional methods, such as classification methods and their requirement for large training samples. The proposed method did not extract the map of buildings using heavy and complex algorithms, which was an important consideration in the discussion of computational cost. Therefore, it can be concluded that the simultaneous use of Radar and optical RS data in the GEE Web-Based platform has a very high potential in distinguishing features and building mapping.