Semi-automatic Classification Research Articles

Software-assisted structured reporting and semi-automated TNM classification for NSCLC staging in a multicenter proof of concept study

ObjectivesIn this multi-center study, we proposed a structured reporting (SR) framework for non-small cell lung cancer (NSCLC) and developed a software-assisted tool to automatically translate image-based findings and annotations into TNM classifications. The aim of this study was to validate the software-assisted SR tool for NSCLC, assess its potential clinical impact in a proof-of-concept study, and evaluate current reporting standards in participating institutions.MethodsA framework for SR and staging of NSCLC was developed in a multi-center collaboration. SR annotations and descriptions were used to generate semi-automated TNM classification. The SR and TNM classification tools were evaluated by nine radiologists on n = 20 representative [18F]FDG PET/CT studies and compared to the free text reporting (FTR) strategy. Results were compared to a multidisciplinary team reference using a generalized linear mixed model (GLMM). Additionally, participants were surveyed on their experience with SR and TNM classification.ResultsOverall, GLMM analysis revealed that readers using SR were 1.707 (CI: 1.137–2.585) times more likely to correctly classify TNM status compared to FTR strategy (p = 0.01) resulting in increased overall TNM correctness in 71.9% (128/178) of cases compared to 62.8% (113/180) FTR. The primary source of variation in classification accuracy was explained by case complexity. Participants rated the potential impact of SR and semi-automated TNM classification as positive across all categories with improved scores after template validation.ConclusionThis multi-center study yielded an effective software-assisted SR framework for NSCLC. The SR and semi-automated classification tool improved TNM classification and were perceived as valuable.Critical relevance statementSoftware-assisted SR provides robust input for semi-automated rule-based TNM classification in non-small-cell lung carcinoma (NSCLC), improves TNM correctness compared to FTR, and was perceived as valuable by radiology physicians.Key PointsSR and TNM classification are underutilized across participating centers for NSCLC staging.Software-assisted SR has emerged as a promising strategy for oncologic assessment.Software-assisted SR facilitates semi-automated TNM classification with improved staging accuracy compared to free-text reports in NSCLC.Graphical

Rapid diagnosis of the geospatial distribution of intertidal macroalgae using large-scale UAVs

Macroalgae have been used as indicators of the health of coastal ecosystems, they function as sinks of CO2 and are essential contributors to primary production. With the increase in anthropogenic activities, it is crucial to assess the impact of such activities on these ecosystems. As traditional surveying techniques, although accurate, are time-consuming and their area coverage is limited, novel techniques are required to monitor the coverage and diversity of intertidal macroalgae. We propose a methodology using the free-source Semi-Automatic Classification Plugin from QGIS to use UAV and multispectral cameras for the spatiotemporal monitoring of intertidal macroalgae. We also compared the performance of six classifiers: Minimum Distance (MD), Maximum Likelihood (ML), Spectral Angle Mapping (SAM), Multi-Layer Perceptron (MLP), Random Forest (RF) and Support Vector Machine (SVM), for three types of macroalgae classification: general, taxonomical groups and species. As proof of concept, an intertidal rocky shore in a marine protected area (NW Spain) was studied for four months. RF and SVM achieved similar results, with both being recommended for the general (OASVM = 97.4±1.7 and OARF = 98.3±1.7) and taxonomical groups (OASVM = 91.6±1.9 and OARF = 89.2±4.5). SVM and ML were found to be more suitable for species classification (OASVM = 77.4±11.4 and OAML = 74.2±9.7). SAM and MLP provided the least performant species classifiers because of the overlap in the macroalgae spectral signatures. The plugin showed limitations when tuning the input parameters of the MLP classifier and did not let to add a validation dataset. Additionally, we present an open-access GIS web application, Alganat 2000 GIS web, to facilitate the monitoring and management of coastal areas. We conclude that the proposed methodology using the SVM or ML classifiers is an effective tool for assessing intertidal macroalgal assemblages. Its easy and rapid implementation is beneficial for researchers who are not very familiar with coding and machine learning frameworks and reduces the time and cost of fieldwork. As future work, we propose the combination of the multispectral bands with topographic and spectral indices and to research the application of deep learning models to the classification of intertidal macroalgae.

Urban Spatial Dynamics and Geo-informatics Prediction of Karachi from 1990–2050 Using Remote Sensing and CA-ANN Simulation

Rapid urbanization significantly impacts land use and land cover (LULC), leading to various socioeconomic and environmental challenges. Effective monitoring and detection of spatial discrepancies are crucial for urban planners and authorities to manage these changes. This study aims to analyze the spatial dynamics of LULC changes and predict future land use patterns. The specific objectives are to classify historical land use from 1990 to 2020, simulate future land use from 2020 to 2050, and interpret the spatial and temporal results. The study utilized remotely sensed images with the semi-automatic classification plugin (SCP) approach for land use classification from 1990 to 2020. Future land use patterns were simulated using the Modules of Land Use Change Evaluation (MOLUSCE)-based Cellular Automata-Artificial Neural Network (CA-ANN) model. The results were then interpreted to comprehend the dynamics of urban expansion. The conclusions direct a significant increase in built-up and grasslands, with a consistent decline in other land use types. From 1990 to 2020, approximately 423.75 km² and 856.97 km² of land were converted into built-up areas and grasslands, respectively. This was accompanied by a decline in rocky bare and bare soil areas, while the proportions of water bodies and mangroves remained steady. Predictions for 2020 to 2050 suggest an additional increase of 561.93 km² in built-up areas, with a progressive decline in other land use classes. The study emphasizes the critical need for spatial planning policies to address challenges arising from rapid urbanization. By analyzing historical land use changes and predicting future patterns this research offers a comprehensive view of urban growth dynamics. The novel application of these techniques provides valuable insights for urban planners to develop informed strategies for managing expansion and mitigating associated socioeconomic and environmental impacts.

Comparative Analysis of Machine Learning Techniques and Data Sources for Dead Tree Detection: What Is the Best Way to Go?

In Central Europe, the extent of bark beetle infestation in spruce stands due to prolonged high temperatures and drought has created large areas of dead trees, which are difficult to monitor by ground surveys. Remote sensing is the only possibility for the assessment of the extent of the dead tree areas. Several options exist for mapping individual dead trees, including different sources and different processing techniques. Satellite images, aerial images, and images from UAVs can be used as sources. Machine and deep learning techniques are included in the processing techniques, although models are often presented without proper realistic validation.This paper compares methods of monitoring dead tree areas using three data sources: multispectral aerial imagery, multispectral PlanetScope satellite imagery, and multispectral Sentinel-2 imagery, as well as two processing methods. The classification methods used are Random Forest (RF) and neural network (NN) in two modalities: pixel- and object-based. In total, 12 combinations are presented. The results were evaluated using two types of reference data: accuracy of model on validation data and accuracy on vector-format semi-automatic classification polygons created by a human evaluator, referred to as real Ground Truth. The aerial imagery was found to have the highest model accuracy, with the CNN model achieving up to 98% with object classification. A higher classification accuracy for satellite imagery was achieved by combining pixel classification and the RF model (87% accuracy for Sentinel-2). For PlanetScope Imagery, the best result was 89%, using a combination of CNN and object-based classifications. A comparison with the Ground Truth showed a decrease in the classification accuracy of the aerial imagery to 89% and the classification accuracy of the satellite imagery to around 70%. In conclusion, aerial imagery is the most effective tool for monitoring bark beetle calamity in terms of precision and accuracy, but satellite imagery has the advantage of fast availability and shorter data processing time, together with larger coverage areas.

Land cover change assessment in Thai Nguyen Province, Vietnam using GIS and remote sensing techniques

The expansion of agricultural production and urbanization has led to the depletion of the global forest ecosystem, resulting in land cover changes (LCCs) that pose threats to the land environment. Therefore, it is imperative to identify these LCCs as an essential step toward resolution and mitigation. This study aims to evaluate temporal LCCs in Thai Nguyen province by utilizing GIS and remote sensing techniques (RST) to analyze high-resolution satellite imagery spanning from 2001 to 2023. LCCs were identified using semi-automatic classification plugin (SCP) techniques with Landsat-8 and Sentinel-2 images in ERVI software. The accuracy of the LCC maps was validated through post-classification comparisons, yielding high precision rates ranging from 87% to 96%. The results indicate a significant reduction in forest cover area, declining by 1018 ha from 226,18 ha in 2001 to 207,22 ha in 2023. This decline primarily occurred in urban areas, highlighting the intensification of urbanization processes. Cultivated and bare land experienced a minor decrease and a substantial decrease from 18,631 ha and 105,000 ha to 16,578 ha and 86,82 ha between 2001 and 2010, followed by a slight increase and a significant increase to 5,069 ha and 24,487 ha (1.4% and 6.9%) during the 2010-2023 period. These findings underscore the escalating risk of diminishing forest cover in the study area.

Evolution of land cover from 1989 to 2023 in the southern region of the Ñeembucú department, Paraguay

This research study, conducted in the southern region of the Ñeembucú Department, Republic of Paraguay, aims to investigate the evolution of land cover from 1989 to 2023. The study area encompasses the districts of Pilar, Humaitá, Paso de Patria, General Díaz, Mayor Martínez, Isla Umbú, Desmochados, Guazu Cuá, Tacuaras, Villalbín, and Laureles. The research was initiated due to concerns regarding the degradation of wetlands, especially since this region is considered unique in the country. In recent years, the southern region of the Ñeembucú Department has undergone numerous human interventions, primarily for productive purposes. The study area, covering a total surface of 6,636.5 km², was demarcated. Subsequently, LANDSAT satellite images were obtained through the U.S. Geological Survey, selected for less than 20% cloud cover. The selected area was then delineated by cropping the images to the specified extent using QGIS 3.32 software and the Semi-Automatic Classification plugin, facilitating the semi-automatic classification of land cover. Significant deforestation, an increase in areas without land cover, urban areas, and sparse vegetation, as well as a decrease in the extent of swamps, forests, grasslands, and prairies were observed.

Geoecological analysis of the state of forest park plantations in the Dnipro city using Landsat satellite images


 
 
 The importance of the number and viability of green areas for creating a com- fortable and safe life for residents of urbanized areas is increasing with the growth of cities, which has not stopped since the first half of the 20th century. Dnipro is a city with progressive air pollution, therefore the effective use of its green areas in this aspect – the selection of tree species, regular assessment of their condition and timely response to changes in vegetative activity – becomes especially relevant. To solve the research task, a combination of on-site and remote methods of assessing the vital state of green areas of the city was chosen. First, spatial and statistical information about the city’s green areas was defined and summarized, and a cartographic and statistical base of raw data was created, potential anthropogenic factors of influence were determined, and reference areas were selected. At the second stage, a systematic survey and assessment of the vital state of the trees of selected areas was carried out for seven years: morphometric indicators of trees of certain species and assessment of the state of their morphological elements (roots, trunk, crown, leaves). According to the degree of damage, trees were classified as: healthy, weakened (damaged), severely weakened; dying and dead. At the third stage of the study, the correlation of the vital state indicator (Ln) and the normalized vegetation index (NDVI) was determined, based on images from Landsat sensors 8 and 9 of the L1 processing level, pre-processing of the images was performed using the SCP (Semi-automatic Classification Plugin) module integrated in QGIS. According to the results of on-site research, it was found that for the period 2015-2021, most parks and squares of the control group belong to the second category (weakened), three to the first (healthy) and one to the third (severely weakened). The main types of damage have been identified as well as species of trees that are most susceptible to damage. The dynamics of vegetation activity indicators reflect the general tendency to increase the vegetation cover of the city for six years. According to the results of the correlation analysis of NDVI and Ln values, a stable dependence between the index of vital state and the normalized vegetation index was revealed. The correlation between the specified indicators for each particular year is direct and varies from 0.5 to 0.9. The vital state of forest park plantations in Dnipro city is at a satisfactory level. The average Ln for the entire city is 65.6%. It was concluded that the exploitation of the gas-absorbing capacity of plants must be carried out in combination with technical measures in order to activate the processes of improving the environment.
 
 

Recognizing surface edges and prediction of terrain deformation by analyzing LANDSAT-8 raster in QGIS- A case study of Coimbatore, India

Convolutional Neural Network (CNN) is one of the deep learning algorithms generally used for image recognition and allocation. These neural networks are developed in multi-layers which reasonably reduces complex dimensions of any input without ruining original information. The satellite images are obtained in .jpg format with potential resolutions. The land usage of the given area is estimated and the objects present in the image are identified using the Canny Edge detection algorithm. It extracts useful data in terms of structures and scales down the size of the data. Raster data in.tif format from LANDSAT-8 is collected over a year. With the Semi-automated classification plugin (SCP) in QGIS, the signatures are created. Signatures are pixelated polygons that are classified to store land attributes. The Normalized indices of vegetation (NDVI), water (NDWI), and built-up (NDBI) are calculated. Land use land cover area was developed. Multi-layer perceptron has numerous hidden layers, and the iterations can be fixed in the MOLUSCE plugin. The land cover for two years, 2020 and 2023 is given along with spatial variables such as precipitation and elevation. The changes in each category of land are identified. In the last three years, the area covered by buildings has increased from 25% to 31%. The area under water bodies had a slight decrease from 1.46% to 1.25%. The land cover for the year 2026 is predicted. From the predictions, it is conclusive that, our research supports the changes between 3 years had not much difference, but above 6 years, it is evident that land will be deformed from most of the vegetation area into built-up.

Microplastics' Shape and Morphology Analysis in the Presence of Natural Organic Matter Using Flow Imaging Microscopy.

Ubiquitous microplastics in urban waters have raised substantial public concern due to their high chemical persistence, accumulative effects, and potential adverse effects on human health. Reliable and standardized methods are urgently needed for the identification and quantification of these emerging environmental pollutants in wastewater treatment plants (WWTPs). In this study, we introduce an innovative rapid approach that employs flow imaging microscopy (FlowCam) to simultaneously identify and quantify microplastics by capturing high-resolution digital images. Real-time image acquisition is followed by semi-automated classification using customized libraries for distinct polyethylene (PE) and polystyrene (PS) microplastics. Subsequently, these images are subjected to further analysis to extract precise morphological details of microplastics, providing insights into their behavior during transport and retention within WWTPs. Of particular significance, a systematic investigation was conducted to explore how the presence of natural organic matter (NOM) in WWTPs affects the accuracy of the FlowCam's measurement outputs for microplastics. It was observed that varying concentrations of NOM induced a more curled shape in microplastics, indicating the necessity of employing pre-treatment procedures to ensure accurate microplastic identification when utilizing the FlowCam. These observations offer valuable new perspectives and potential solutions for designing appropriate treatment technologies for removing microplastics within WWTPs.

Applying High-Resolution Satellite and UAS Imagery for Detecting Coldwater Inputs in Temperate Streams of the Iowa Driftless Region

Coldwater streams are crucial habitats for many biota including Salmonidae and Cottidae species that are unable to tolerate warmer water temperatures. Accurate classification of coldwater streams is essential for their conservation, restoration, and management, especially in light of increasing human disturbance and climate change. Coldwater streams receive cooler groundwater inputs and, as a result, typically remain ice-free during the winter. Based on this empirical thermal evidence, we examined the potential of very high-resolution (VHR) satellite and uncrewed aerial system (UAS) imagery to (i) detect coldwater streams using semi-automatic classification versus visual interpretation approaches, (ii) examine the physical factors that contribute to inaccuracies in detecting coldwater habitats, and (iii) use the results to identify inaccuracies in existing thermal stream classification datasets and recommend coverage updates. Due to complex site conditions, semi-automated classification was time consuming and produced low mapping accuracy, while visual interpretation produced better results. VHR imagery detected only the highest quality coldwater streams while lower quality streams that still met the thermal and biological criteria to be classified as coldwater remained undetected. Complex stream and site variables (narrow stream width, canopy cover, terrain shadow, stream covered by ice and drifting snow), image quality (spatial resolution, solar elevation angle), and environmental conditions (ambient temperature prior to image acquisition) make coldwater detection challenging; however, UAS imagery is uniquely suited for mapping very narrow streams and can bridge the gap between field data and satellite imagery. Field-collected water temperatures and stream habitat and fish community inventories may be necessary to overcome these challenges and allow validation of remote sensing results. We detected >30 km of coldwater streams that are currently misclassified as warmwater. Overall, visual interpretation of VHR imagery it is a relatively quick and inexpensive approach to detect the location and extent of coldwater stream resources and could be used to develop field monitoring programs to confirm location and extent of coldwater aquatic resources.

Hybrid system model for wind abrasion segmentation using semi-automatic classification of remote sensing multispectral areas

Wind abrasion, caused by particles transported by strong winds impacting on structures, can lead to their degradation. Although this phenomenon has hardly been studied in this context, it is becoming increasingly important due to new trends in infrastructure location, especially in renewable energy terms. Metallic structures are particularly vulnerable to degradation by the action of windblown sand particles. However, characterising such secluded sites is complicated, and remote sensing systems and satellite information become crucial. The objective of this research is to identify and delineate the geographic areas that are vulnerable to this phenomenon by employing a hybrid model with historical data and the semi-automatic classification of multispectral satellite images. The model is based on critical variables identified by the scientific community and case studies documented in the literature. The methodology used for the study consists of four phases, including creating a scientifically robust database, downloading and managing satellite and historical long-term information, segmenting the regions of interest, and modelling using supervised classification techniques. The proposed algorithm shows very accurate results (R2 = 0.9922) and the overall system approach is presented as a useful and generalizable method to address this problem, increasing the existing knowledge on material wear by particle action, and contributing to optimizing the initial design of resilient structures.

The Conceptual Grounding of Overtourism and Overtourism-Driven Change: Olympos Case

Heritage tourism destinations (HTDs) in rural places host not only tourists and tourism service facilities but also negative consequences such as urbanization and overtourism-based spatial transformation on the social, economic, and physical landscape. This paper aims to investigate the ontological groundings of a rehashed overtourism phenomenon by employing an integrative reviewing method of related literature with a focus on rural HTDs and reveal the landscape change by extracting the spatial transformation from satellite images and historical orthophotos of Olympos/Turkey by using semi-automatic classification analysis in QGIS. The findings implicate that the indicators of overtourism can be grounded on the latest levels of the Creative Destruction Model, the Vicious Circle, and the Tourismification approaches, however, still there is a need for reconceptualization of the phenomenon. Moreover, the findings showed that the modus operandi of overtourism-based spatial expansion of tourism service units in protected areas follows a path through the gaps between the two inverse philosophies of protection and use which is critical for stage changes in the evolution process of HTD.

Mapping of Inland Excess Water Using Geographical Information System and High-Resolution Satellite Images: A Case Study of SREM, Serbia

Abstract Extreme hydrological events, such as floods and droughts, are becoming more frequent as a result of climate change, leading to negative impacts on various economic sectors. The Pannonian-Carpathian Basin is particularly affected by the increasing frequency of hazardous hydrological events. Agricultural production, which is a highly significant economic sector in the region, is particularly vulnerable to these unfavourable climatic conditions. Changes in precipitation patterns and soil moisture levels can lead to reduced crop yields, while floods can pollute water sources and erode fertile soil. Mapping of Inland Excess Water (IEW), also known as ponding water or waterlogged areas, is crucial for informed decision-making, damage compensation, risk management, and future prevention planning. Remote sensing technology and machine learning have been demonstrated to be valuable tools for the mapping of IEW. The 2014 floods in Southeastern and Central Europe serve as a reminder of the importance of effective flood risk management. This study used a Geographical Information System (GIS) and a Semi-automated Classification Processing (SCP) tool to process high-resolution RapidEye satellite images from the 2014 floods in the Srem region of Serbia. The Spectral Angle Mapping (SAM) classification model was used to produce a map of IEW. The SAM model achieved an overall accuracy of 92.68 %. The study found that IEW affected approximately 2.90 % or 99.59 km² of the territory in Srem. The obtained maps can be used by responsible water management agencies to prevent and control excessive inland water.

Semi-automated classification of layered rock slopes using digital elevation model and geological map

Abstract Layered rock slopes are the most widely distributed slopes with the simplest structure. The classification of layered rock slopes is the basis for correctly analyzing their deformation and failure mechanisms, evaluating their stability, and adopting reasonable support methods. It is also one of the essential indicators to support the evaluation of urban and rural construction suitability and the assessment of landslide hazards. However, the present-day classification methods for layered rock slopes are not sufficiently automated. In the application process of these methods, a lot of manual intervention is still needed, and sufficient strata orientation data obtained through field surveys is required, which is not effective for large-scale applications and involves high subjectivity. Thus, this study proposes a semi-automated classification method for layered rock slopes based on digital elevation model (DEM) and geological maps, which greatly reduces human intervention. On the basis of slope unit division, the method extracts structural information of slopes using DEM and geological maps and classifies slopes according to their structural characteristics. An experiment has been carried out in the northern region of Mount Lu in Jiangxi Province, and the results demonstrate the effectiveness of this semi-automated classification method. Compared to the existing manual or semi-automated classification methods, the method proposed in this article is objective and highly automated, which can meet the requirements of classification of layered rock slopes over large areas, even in the case of sparse measured orientation data.

Towards standardized reporting and failure classification of safety equipment: Semi-automated classification of failure data for safety equipment in the operating phase

Safety instrumented systems (SISs) are installed on process plants to protect against undesired events like e.g., gas leakage and overpressure. A SIS has reliability requirements that are determined during design, and conformance to these requirements should be verified during operation. It is therefore important that all SIS failures are recorded and classified according to their impact on the SIS reliability. Failures of SIS equipment classified as dangerous undetected are of particular interest because they are dormant (undetected) and will prevent the execution of the safety function (dangerous). Analysis of the failure mode and detection method is essential when deciding if a failure is dangerous and undetected. Such information is often provided as unstructured text in notifications registered into the maintenance management system. Therefore, the work of classifying failures requires considerable manual effort in reading and analyzing the texts. Approaches within natural language processing, like technical language processing, have the potential to be deployed more actively for this purpose. However, successful adoption relies on groundwork where classification rules are derived from international standards and commonly agreed industry practice. This paper presents a semi-automated process that incorporates classification rules and gives examples that indicate some of the capabilities of technical language processing for failure classification. The paper also elaborates on how the work relates to Industry 4.0 in creating digital representations to monitor the performance of safety instrumented systems. This work has been carried out as part of the APOS project (Automated process for follow-up of safety instrumented systems). The APOS project has developed knowledge and specifications that simplify and automate the design and operation of safety equipment and investigated how the failure classification process can be made more efficient.

Semi-automatic classification of digital heritage on the Aïoli open source 2D/3D annotation platform via machine learning and deep learning

Today, digital tools represent a privileged means of describing and understanding of cultural heritage. However, existing hybrid methods for semantic annotation on digital models are mostly manual and require considerable human intervention. This may cause data dispersion or data alteration issues, preventing the digital collection of multidisciplinary heritage information in collaborative valorization and conservation projects. Tackling this problem, this work presents three different methods for more automated 2D/3D data annotation, making the most of Machine Learning or Deep Learning and of the collaborative, web-based annotation platform Aïoli. The proposed approaches are designed to assist heritage experts in the hybrid annotation of architectural objects, e.g., to identify building components, degradation patterns, mapping of alterations, materials etc. and to share and access 2D/3D information via the web. The work aims at providing public and private stakeholders in charge of restoration and conservation activities with a set of more automated annotation tools, allowing to: i) preserve data in the transition to different representation types; ii) propagate analytically relevant information via digital models and/or images; iii) share heritage-related data on the web with an open source strategy. The validated workflows sketch the outlines of a dedicated and user-friendly toolbar for semi-automatic annotation, directly integrated into the Aïoli user interface, where automated annotation tools are readily available to the community of heritage professionals to properly support documentation and restoration activities.

Impact of land use and land cover dynamics on urban heat island in the Sunyani Municipality using satellite remote sensing

The rapid urbanization coupled with tremendous growth in population and built-up areas in the cities has led to issues associated with global climate change impacts such as urban heat islands. This study is concerned with the spatial impact of the dynamics of urbanisation on urban heat island (UHI) in the Sunyani Municipality using Satellite Remote Sensing techniques. Sunyani municipal is a growing area in Ghana with increasing population and urbanisation. Many people in Ghana are moving to settle in Sunyani for different purposes. In this process, vegetated areas have been converted to urban areas and this has led to a reduction in the vegetation cover and increased the urban temperature in the municipality. Therefore, there is a need for a study on how the dynamics in the land use and land cover (LULC) in the municipal are contributing to urban heat. The main aim of this research is to identify the impact of LULC dynamics on UHI using multi-temporal Landsat images from 2001 to 2021 with a 5-year epoch in the Sunyani Municipal. Landsat images acquired over Sunyani and its environs for the period from 2001 to 2021 with a 5-year epoch, were download freely from the USGS Earth Explorer website. These images were pre-processed in QGIS 3.16 using the Semi-automatic classification plugin (SCP). The pre-processed images were classified using the Random Forest (RF) algorithm. The supervised classification was done using the semi-automatic classification plugin in QGIS 3.16. This study considered two (2) classes (urban and vegetation). The thermal infrared bands of the Landsat images were used to compute the land surface temperature (LST) for each year, starting from 2001 up to 2021, at a 5-year interval. The standard deviation and mean LST were extracted and used to extract all the UHIs in the municipality. Regression analyses were used to determine the impact of LULC dynamics on LST and UHIs in the Sunyani Municipality. The analysis yielded a strong impact with an R-squared value of 75% and 77% for UHI and LST respectively. The vegetated covers yielded a strong negative correlation with UHIs while Urbanisation yielded a strong positive correlation with the intensity of UHIs. On the other hand, Vegetation reduced the intensity of UHIs in the municipality. This gives a clear understanding of how dynamics in urbanisation and vegetation are affecting UHIs in the Sunyani Municipality. A Linear model was developed for LULC dynamics, LST and UHIs in the Sunyani municipality. It was concluded that LULC dynamics have a strong impact on UHI. It is recommended that the environmental Agencies in the Municipality should ensure the protection of the vegetated areas as well as improve the green areas in the Municipality to minimize urban heat.

POINT CLOUD CLASSIFICATION OF AN URBAN ENVIRONMENT USING A SEMI-AUTOMATIC APPROACH

Abstract. In this contribution a point cloud classification in an urban context has been presented. The aim of the work is to test a semi-automatic classification approach and to verify its usefulness in the scan-to BIM process, and to validate how much it is straightforward for the definition of different point cloud LODs. The work methodology is structured in three phases. The first concerns data acquisition and processing through geomatic instruments and methodologies that guarantee a complete and expeditious survey such as ground-based MMS and UAV for aerial photogrammetry. The second phase concerns the testing of an online software that per-forms point cloud classifications through AI algorithms. The system allows either to use standard classifiers that are already available, or to create a customizable catalogue of the different classes that one wants to attribute to the urban scene. Following the automatic classification process, where all objects have been identified, manual corrections can be made to improve the classification of objects into specific classes. The third step is object detection and extraction. Here, the relationship between automatic classification, point cloud density, object identification and the various degrees of LOD definition was explored. The higher the LOD, the greater the number of objects that can be identified, particularly those elements related to street furniture and urban facilities. Once these objects have been classified, it is then possible to extract them in interoperable format. This allows such data to be managed and shared through BIM platform.

Improving estimation of water soil erosion by introducing lithological formation for environmental remediation

Soil erosion is a serious and complex environmental problem worldwide, especially in the centre west of Tunisia. Whereas the construction of hill reservoirs is part of the soil and water conservation strategy, many of these have a siltation problem. Dhkekira is one of the smallest watersheds in central Tunisia whose most lithological formation consists of materials that are quite susceptible to water erosion. Due to the lack of low-scale lithological data, digital IR aerial photos with 2 m spatial resolution were considered. A semi-automatic classification of aerial photos, based on the image's textural indices is developed. The lithologic map extracted from aerial photos was used as input for ANSWERS-2000 water erosion model. Results obtained indicate first, with the semi-automatic classification of the mean and standard deviation of the thumbnail histograms that image output could help to give an idea about the existence of some surface lithological formation. The model applied to Dhkekira watershed showed that the spatial difference in water erosion was not caused only by land cover and slope, but also by lithological formation. The percentage of each lithological formation in sediment yield at the Dhkekira hill reservoir was estimated to be 69% sediment yield from Pleistocene and 19.7% from Lutetian-Priabonian.

Automated Subject Identification using the Universal Decimal Classification: The ANN Approach

Universal Decimal Classification (UDC) is a popular controlled vocabulary that is used to represent subjects of documents. Text categorization determines a text's category, as evident from the notation-text label format of the Universal Decimal Classification. With the help of machine learning techniques and the Universal Decimal Classification (UDC), the present work aims to develop an end-user (library professional) based recommender system for automatically classifying documents using the UDC scheme. The proposed work is conceived for determining and constructing a complex class number using the syntax of Universal Decimal Classification (UDC). A corpus of documents classified with the UDC scheme is used as a training dataset. The classification of the documents is done with human mediation having proficiency in classificatory approaches. The BERT model and the KNIME software are used for the study. This study uses the classified dataset to fine-tune the pre-trained BERT model to construct the semi-automatic classification model. The results show that the model is constructed with high accuracy and Area Under Curve (AUC) value, although the prediction represented a low accuracy rate. This study reflected that if the model is explicitly trained by annotating each concept and if the full licensed version of UDC class numbers becomes available, there is a greater potency of developing an automated, freely faceted classification scheme for practical use.