Type Mapping Research Articles

Remote Sensing Classification of Temperate Grassland in Eurasia Based on Normalized Difference Vegetation Index (NDVI) Time-Series Data

The Eurasian temperate grassland is the largest temperate grassland ecosystem and vegetation transition zone globally. The spatiotemporal distribution and changes of grassland types are vital for grassland monitoring and management. However, there is currently a lack of a unified classification method and standard distribution map of Eurasian temperate grassland types. The Normalized Difference Vegetation Index (NDVI) from remote sensing data is commonly used in grassland monitoring. In this paper, the Accumulated Rate of NDVI Change Index (ARNCI) was proposed to characterize the annual NDVI trend of different temperate grassland types, and four transitional categories were introduced to account for the overlap between them. Based on survey data on the distribution of Eurasian temperate grassland types in the 1980s, the study area was divided into three sub-regions: Northern China, Central Asia, and Mongolia. Regionally, pixel-based ARNCI maps in the 1980s and 1990s were successfully calculated from using NOAA’s AVHRR NDVI time-series products. The ARNCI classification thresholds for different sub-regions were determined, and classification experiments and validation were conducted for each sub-region. The overall accuracies of grasslands types classification for Northern China, Central Asia, and Mongolia in the 1980s were 75.3%, 64.2%, and 84.6%, respectively, which demonstrated that there were variations in classification accuracy in the three sub-regions, and the overall performance was favorable. Finally, distribution maps of Eurasian temperate grassland types in the 1980s and 1990s were obtained, and the spatiotemporal changes of grassland types were analyzed and discussed. The ARNCI method is simple to operate and easy to obtain data, and it can be conveniently used in grassland type classification. The maps firstly address the lack of remote sensing classification maps of Eurasian temperate grassland types, and provide a promising tool for monitoring grassland degradation, management, and utilization.

On a multivalued prescribed mean curvature problem and inclusions defined on dual spaces

This article addresses two main objectives. First, it establishes a functional analytic framework and presents existence results for a quasilinear inclusion describing a prescribed mean curvature problem with homogeneous Dirichlet boundary conditions, involving a multivalued lower order term. The formulation of the problem is done in the space of functions with bounded variation. The second objective is to introduce a general existence theory for inclusions defined on nonreflexive Banach spaces, which is specifically applicable to the aforementioned prescribed mean curvature problem. This problem can be formulated as a multivalued variational inequality in the space of functions with bounded variation, which, under suitable conditions, is equivalent to an inclusion involving a maximal monotone mapping of type (D) and a generalized pseudomonotone mapping. We prove an abstract existence theorem for inclusions of this form, under some coercivity conditions involving both the maximal monotone and the generalized pseudomonotone mappings.

A marmoset brain cell census reveals regional specialization of cellular identities.

The mammalian brain is composed of many brain structures, each with its own ontogenetic and developmental history. We used single-nucleus RNA sequencing to sample over 2.4 million brain cells across 18 locations in the common marmoset, a New World monkey primed for genetic engineering, and examined gene expression patterns of cell types within and across brain structures. The adult transcriptomic identity of most neuronal types is shaped more by developmental origin than by neurotransmitter signaling repertoire. Quantitative mapping of GABAergic types with single-molecule FISH (smFISH) reveals that interneurons in the striatum and neocortex follow distinct spatial principles, and that lateral prefrontal and other higher-order cortical association areas are distinguished by high proportions of VIP+ neurons. We use cell type-specific enhancers to drive AAV-GFP and reconstruct the morphologies of molecularly resolved interneuron types in neocortex and striatum. Our analyses highlight how lineage, local context, and functional class contribute to the transcriptional identity and biodistribution of primate brain cell types.

Fuel Type Mapping Using a CNN-Based Remote Sensing Approach: A Case Study in Sardinia

Accurate fuel mapping is crucial for effectively determining wildfire risk and implementing management strategies. The primary challenge in fuel type mapping lies in the need to develop accurate and efficient methods for identifying and categorizing the various combustible materials present in an area, often on a large scale. In response to this need, this paper presents a comprehensive approach that combines remote sensing data and Convolutional Neural Network (CNN) to discriminate between fire behavior fuel models. In particular, a CNN-based classification approach that leverages Sentinel-2 imagery is exploited to accurately classify fuel types into seven preliminary main classes (broadleaf, conifers, shrubs, grass, bare soil, urban areas, and water bodies). To further refine the fuel mapping results, subclasses were generated from the seven principles by using biomass and bioclimatic maps. These additional maps provide complementary information about vegetation density and climatic conditions, respectively. By incorporating this information, we align our fuel type classification with the widely used Standard Scott and Burgan (2005) fuel classification system. The results are highly promising, showcasing excellent CNN training performance with all three metrics—accuracy, recall, and F1 score—achieving an impressive 0.99%. Notably, the network exhibits exceptional accuracy in a test case conducted in the southern region of Sardinia, successfully identifying Burnable classes in previously unseen pixels: broadleaf at 0.99%, conifer at 0.79%, shrub at 0.76%, and grass at 0.84%. The proposed approach presents a valuable tool for enhancing fire management, contributing to more effective wildfire prevention and mitigation efforts. Thus, this tool could be leveraged by fire management agencies, policymakers, and researchers to improve the determination of wildfire risk and management.

Pol-InSAR-Island - A benchmark dataset for multi-frequency Pol-InSAR data land cover classification

This paper presents Pol-InSAR-Island, the first publicly available multi-frequency Polarimetric Interferometric Synthetic Aperture Radar (Pol-InSAR) dataset labeled with detailed land cover classes, which serves as a challenging benchmark dataset for land cover classification. In recent years, machine learning has become a powerful tool for remote sensing image analysis. While there are numerous large-scale benchmark datasets for training and evaluating machine learning models for the analysis of optical data, the availability of labeled SAR or, more specifically, Pol-InSAR data is very limited. The lack of labeled data for training, as well as for testing and comparing different approaches, hinders the rapid development of machine learning algorithms for Pol-InSAR image analysis. The Pol-InSAR-Island benchmark dataset presented in this paper aims to fill this gap. The dataset consists of Pol-InSAR data acquired in S- and L-band by DLR's airborne F-SAR system over the East Frisian island Baltrum. The interferometric image pairs are the result of a repeat-pass measurement with a time offset of several minutes. The image data are given as 6 × 6 coherency matrices in ground range on a 1 m × 1m grid. Pixel-accurate class labels, consisting of 12 different land cover classes, are generated in a semi-automatic process based on an existing biotope type map and visual interpretation of SAR and optical images. Fixed training and test subsets are defined to ensure the comparability of different approaches trained and tested prospectively on the Pol-InSAR-Island dataset. In addition to the dataset, results of supervised Wishart and Random Forest classifiers that achieve mean Intersection-over-Union scores between 24% and 67% are provided to serve as a baseline for future work. The dataset is provided via KITopenData: https://doi.org/10.35097/1700.

Remote Sensing Techniques for Identifying Flood Prone Areas Using the Vegetation Index, Water Index, and Simple Additive Weighting Methods for Mekarsari Areas, Cianjur Regency, West Java

Flood is a natural disaster with an intensity that often occurs in cities on a scale with excess water discharge in some areas. This research was conducted to determine the risk of flooding in Mekarsari Village and its surroundings, Cianjur Regency, West Java. (According to InaRisk, 2022) Risk and danger of flooding Moderate – High. This study uses data for the entire year 2022. With several parameters used in this study including land cover, rainfall, soil type, slope, NDVI, NDWI and SAVI. Simple Additive Weighting or weighting for each parameter with different results, according to the classification of each of these parameters. The results of this weighting will become a Flood Hazard Risk Map which was previously the result of overlay and intersection of land cover map, rainfall map, soil type map, slope map, Normalized Difference Vegetation Index map, Normalized Difference Water Index map and Soil Adjusted Vegetation Index map. On the Flood Disaster Risk Map there are five categories of vulnerability levels Not Hazardous, Less Hazardous, Moderate, Hazardous and Very Hazardous. The cause of the risk of flooding in the study area is that most of the land use is diverted to rice fields, plantations, dry agricultural land and settlements so that the lack of vacant land causes low absorption of rainwater and also steep slopes and high elevations which also cause rainwater to be stored in the lowlands. settlements) with moderate to high rainfall intensity every month. There is also a lack of education for the community to mitigate the risk of flood disasters so that waste management and waterways in settlements are inadequate and well managed.

Extraction, labelling, clustering, and semantic mapping of segments from clinical notes.

This work is motivated by the scarcity of tools for accurate, unsupervised information extraction from unstructured clinical notes in computationally underrepresented languages, such as Czech. We introduce a stepping stone to a broad array of downstream tasks such as summarisation or integration of individual patient records, extraction of structured information for national cancer registry reporting or building of semi-structured semantic patient representations that can be used for computing patient embeddings. More specifically, we present a method for unsupervised extraction of semantically-labelled textual segments from clinical notes and test it out on a dataset of Czech breast cancer patients, provided by Masaryk Memorial Cancer Institute (the largest Czech hospital specialising exclusively in oncology). Our goal was to extract, classify (i.e. label) and cluster segments of the free-text notes that correspond to specific clinical features (e.g., family background, comorbidities or toxicities). Finally, we propose a tool for computer-assisted semantic mapping of segment types to pre-defined ontologies and validate it on a downstream task of category-specific patient similarity. The presented results demonstrate the practical relevance of the proposed approach for building more sophisticated extraction and analytical pipelines deployed on Czech clinical notes.

EVALUASI FUNGSI LINGKUNGAN EMBUNG BERBASIS SISTEM INFORMASI GEOGRAFI: KASUS EMBUNG KRAPYAK DAN TONOGORO DI KULON PROGO, DAERAH ISTIMEWA YOGYAKARTA (Environment Role Evaluation of Retention Basins Using Geographic Information Systems: Case of Krapyak and Tonogoro in Kulon Progo, Special Region of Yogyakarta)

Environmentally, retention basins are used for soil and water conservation but some of them experience drought. The location of the retention basin is fundamental in fulfilling environmental functions. This study aimed to apply GIS for evaluating the environmental functions of retention basins in a certain area. Tonogoro and Krapyak retention basins lay on Kulon Progo Regency, Yogyakarta are the cases of this study. The Krapyak is located at the outlet of catchment area, while the Tonogoro is built on the hilltop. Environmental functions were assessed based on the potential drought and dynamics of carrying capacity due to sedimentation. The materials used were detailed design documents, Google Earth images, DEMNAS images, 1:25,000 scale Indonesian landform maps, rainfall data, and soil type maps. The initial reservoir capacity was calculated based on depth using DEMNAS data. The recent reservoir capacity was measured using a sounding method with a Garmin 520s Echo sounder, Depth Sounder, and Roll meter. All data was integrated in GIS to be presented visually in the form of a reservoir depth profile that reflects the dynamics of the reservoir capacity. The potential of surface runoff volume was calculated using SCS-CN method. This study found that the Tonogoro retention basin experienced a water shortage because the potential volume of runoff water was only able to supply 5% of the reservoir’s capacity. The accumulated sediment yield in the Krapyak retention basins was 3,498 m3 or 12.2% of the total capacity of a sedimentation rate of 699.6 m3 per year.

Applying machine learning classifiers to Sentinel-2 imagery for early identification of cotton fields to advance boll weevil eradication

Early identification of cotton fields is critical for boll weevil eradication programs to establish field traps for monitoring weevil populations. However, most studies on crop type mapping have focused on late- or post-season crop classification. This study proposed and evaluated a field-level classification method that integrated field boundary data and early-season Sentinel-2 imagery to identify cotton fields 20–60 days after planting. Three cloud-free images were selected from May and June 2020, covering a 110-square-km study area in Texas. To achieve field-level classification, the boundaries of all 320 fields in the study area were used, and 62 of them were selected as training fields. Three machine learning classifiers - random forest (RF), support vector machine (SVM), and k-nearest neighbor (KNN) - were applied to the four 10-m bands for pixel-based classification and to segmented images by mean shift segmentation for object-based classification to determine the optimal classifier. To assess the impact of additional bands and vegetation indices (VIs) on classification accuracy, the six 20-m bands were combined with the 10-m bands to create 10-band images with 10-m resolution, and six vegetation indices (VIs) were derived and combined with the images. Recursive feature selection was used to identify optimal band combinations and principal component (PC) analysis was applied to reduce the dimensionality of the images. Based on accuracy measures derived from 2160 error matrices for all classification maps, RF provided the highest accuracy with faster processing times for the two May dates, while SVM and KNN exhibited slightly higher accuracy for June 10. The integration of field boundary data significantly improved classification accuracy for all classification maps. The overall accuracy (OA) of the optimal RF classification maps for the 4-band images increased from 71.4 %, 76.3 %, and 82.6 % to 85.8 %, 90.2 %, and 92.6 % (a net increase of 14.4 %, 13.9 %, and 10.0 %) for May 1, May 6, and June 10, respectively. Moreover, by including the six 20-m bands, the OA further increased by 4.8 %, 2.9 %, and 3.0 % for the respective dates. Similarly, the F-score for cotton increased from 0.801, 0.845, and 0.882 to 0.936, 0.947, and 0.976 for the three respective dates. In contrast, the inclusion of VIs had minimal effect on classification accuracy. The optimal bands and the PCs provided better classification results and were used for image classification to reduce and optimize input feature to the classifiers. These results demonstrate that combining RF with 10-band Sentinel-2 imagery and field boundary data is an effective method for early identification of cotton fields. This study provides practical techniques for boll weevil eradication program managers to identify cotton fields over large geographic areas during early growth stages.

Using machine learning to generate an open-access cropland map from satellite images time series in the Indian Himalayan region

Crop maps are crucial for agricultural monitoring and food management and can additionally support domain-specific applications, such as setting cold supply chain infrastructure in developing countries. Machine learning (ML) models, combined with freely-available satellite imagery, can be used to produce cost-effective and high spatial-resolution crop maps. However, accessing ground truth data for supervised learning is especially challenging in developing countries due to factors such as smallholding and fragmented geography, which often results in a lack of crop type maps or even reliable cropland maps. Our area of interest for this study lies in Himachal Pradesh, India, where we aim at producing an open-access binary cropland map at 10-m resolution for the Kullu, Shimla, and Mandi districts. To this end, we developed an ML pipeline that relies on Sentinel-2 satellite images time series. We investigated two pixel-based supervised classifiers, support vector machines (SVM) and random forest (RF), which are used to classify per-pixel time series for binary cropland mapping. The ground truth data used for training, validation and testing was manually annotated from a combination of field survey reference points and visual interpretation of very high resolution (VHR) imagery. We trained and validated the models via spatial cross-validation to account for local spatial autocorrelation and improve the generalization capability of the model. We tested the model on hold out test sets of each district, achieving an average accuracy for the RF (our best model) of 87%. We noticed NIR band at the early and late stage of the apple harvest season (main crop in the region) to be of critical importance for the model. Finally, we used this model to generate a cropland map for three districts of Himachal Pradesh, spanning 14,600 km2, which improves the resolution and quality of existing public maps, and made the code open-source.

A system for concurrent on-the-go soil apparent electrical conductivity and gamma-ray sensing in micro-irrigated orchards

On-the-go soil apparent electrical conductivity (ECa) sensors are great tools for mapping and monitoring soil properties such as water content, texture, and salinity. ECa maps and surveys are most useful and reliable when obtained in uniformly wet fields. However, soil moisture in micro-irrigated (e.g., drip or micro-sprinklers) orchards is typically non-uniform, with moist soil along tree and irrigation lines, and dry soil between tree rows. We developed a mobile platform and data post-processing algorithm to facilitate geospatial ECa measurements along or near driplines. Gamma-ray (γ-ray) spectrometry is commonly used for clay content and type mapping. Fusion between ECa and γ-ray is often reported to increase the accuracy of field-scale soil maps. However, contrarily to ECa, γ-ray spectrometry is best suited for sensing soils in dry conditions. Micro-irrigated orchards are ideal environments for the combined application of these two sensor technologies. The fusion of topsoil (top 0.5 m) ECa (measured along the driplines) and γ-ray total counts (TC) (measured between the tree rows) data was tested at a 0.4-ha sandy loam citrus orchard in Southern California. Here, we discuss sensor data acquisition, data processing, sensor-directed sampling scheme delineation, and characterization of field-scale soil particle size fraction (0–0.4 m soil profile) spatial variability. Pearson correlation coefficients between sand and silt content with both ECa and TC were significant (p < 0.05). A principal component analysis biplot suggested strong positive relationship with TC and clay content. Backwards stepwise multiple linear regression predicted sand content using TC, elevation, and spatial coordinates as explanatory variables with mean absolute error (MAE) of 3.06 %. Silt content was predicted (MAE=1.55 %) using ECa, elevation, and spatial coordinates. The development of this platform enables better characterization of soil properties in micro-irrigated orchard systems using on-the-go sensing technology.

Agricultural Land Cover Mapping through Two Deep Learning Models in the Framework of EU’s CAP Activities Using Sentinel-2 Multitemporal Imagery

The images of the Sentinel-2 constellation can help the verification process of farmers’ declarations, providing, among other things, accurate spatial explicit maps of the agricultural land cover. The aim of the study is to design, develop, and evaluate two deep learning (DL) architectures tailored for agricultural land cover and crop type mapping. The focus is on a detailed class scheme encompassing fifteen distinct classes, utilizing Sentinel-2 imagery acquired on a monthly basis throughout the year. The study’s geographical scope covers a diverse rural area in North Greece, situated within southeast Europe. These architectures are a Temporal Convolutional Neural Network (CNN) and a combination of a Recurrent and a 2D Convolutional Neural Network (R-CNN), and their accuracy is compared to the well-established Random Forest (RF) machine learning algorithm. The comparative approach is not restricted to simply presenting the results given by classification metrics, but it also assesses the uncertainty of the classification results using an entropy measure and the spatial distribution of the classification errors. Furthermore, the issue of sampling strategy for the extraction of the training set is highlighted, targeting the efficient handling of both the imbalance of the dataset and the spectral variability of instances among classes. The two developed deep learning architectures performed equally well, presenting an overall accuracy of 90.13% (Temporal CNN) and 90.18% (R-CNN), higher than the 86.31% overall accuracy of the RF approach. Finally, the Temporal CNN method presented a lower entropy value (6.63%), compared both to R-CNN (7.76%) and RF (28.94%) methods, indicating that both DL approaches should be considered for developing operational EO processing workflows.

Changes in the Occurrence of Five Invasive Plant Species in Different Ecosystem Types between 2009–2018 in Hungary

Modelling and analysis of spatiotemporal characteristics of plant invasion can help in mapping and predicting the spread of invasive plants. The aim of our research was to investigate the spatiotemporal variability of five common invasive plant species (Ailanthus altissima, Asclepias syriaca, Elaeagnus angustifolia, Robinia pseudoacacia, and Solidago spp.) within different land cover (ecosystem)-type categories. The basis of the study was the National Geospatial Database of Invasive Plants (NGDIP) of Hungary, and the ecosystem types of the Ecosystem Map of Hungary (EMH). The GIS-based analysis of the detailed occurrence database of the invasive species (NGDIP) and the thematic land-cover (ecosystem)-type maps (EMH) examined allow us to answer the question of in which habitat types the occurrence and distribution of the given invasive plant has stagnated, decreased, or increased between 2006 and 2018. We developed a methodology with relevant data sources and demonstrated invasion variation, which can be used for future management planning and invasive biology research. Our results show that Asclepias syriaca and Robinia pseudoacacia are increasingly threatening grasslands and are also spreading more intensively in complex cultivated areas. The occurrences of Ailanthus altissima and Asclepias syriaca are declining in built-up areas due to the increasingly extreme environmental conditions of cities or modified urban planning. The spread of Solidago spp. is increasingly common in wetlands, threatening the biodiversity of floodplain (riparian) vegetation.

Management Implications at a Protected, Peri-Urban, Salt Lake Ecosystem: The Case of Larnaca’s Salt Lakes (Cyprus)

Larnaca’s salt lakes constitute one of the most important protected areas in Cyprus and in the southeast Mediterranean. It is in continuity with the urban area of Larnaca city, being the major green infrastructure in the area, supporting (among others) biodiversity, recreation, culture, and food production. The present study aims to provide an overview of the area’s ecosystem types and their potential to provide ecosystem services, focusing on assessing the water ecosystem condition and drafting the relevant baseline accounts. The results of the study include an ecosystem type map of the area, followed by potential supply maps of ecosystem services, for the three major categories (i.e., provisioning, regulating and maintenance, and cultural) and the estimation and assessment of ecosystem condition variables for wetlands, as proposed by the EU-wide methodology to map and assess the ecosystem condition, in line with the System of Environmental Economic Accounting—Ecosystem Accounting (SEEA EA). A critical exceedance (over 10%) of the imperviousness and the increase in the soil moisture deficit within the wider salt lakes area have been identified and can constitute useful variables associated with the anthropogenic, climatic, and other effects on the condition of the salt lakes. Discussion for integrating this information to existing knowledge is provided toward supporting scientifically informed decision and policymaking for the sustainable development of protected areas.

Improving high school students’ Physics learning outcomes using mind maps

Based on the interview with accelerated second-grade senior high school students, most of them had difficulty remembering many physics subject matter in a short time, such as when facing a final exam. Likewise, from the literature study, it was found that there was no presentation of physics material that was easy to remember quickly in a short time. One solution idea that can be given is to use a Mind Map. This study aims to improve high school students learning outcomes in Physics by using Mind Maps and finding types of Mind Maps that are easy to learn. The research method used is a class action research method consisting of three cycles. Before using mind maps, the average physics learning outcomes were obtained at 48.3. In the first cycle using mind maps of the cobweb type, the average result obtained was 60.54, an increase of 25.34%. While the second cycle using tree root mind maps, the average obtained was 71.5, experiencing an increase of 18.1% from cycle 1. Meanwhile, cycle III using tree root mind map types obtained an average student physics learning outcome of 86.5, increased results by 20.97%. From the results of this study, it can be concluded that the use of mind maps in physics learning can improve student learning outcomes. In comparison, the type of mind map that has the greatest influence is the type of tree roots. In other words, tree root-type mind maps can present physics material that is more easily understood by students.

Sedimentation analysis using SWAT model (soil and water assessment tool) in Mamasa Sub-Watershed

Soil fertility and quality of agricultural land have decreased due to erosion and sedimentation in the sub-watershed upstream area. This decrease in soil fertility is due to the loss of NPK nutrients in the topsoil. Erosion and sedimentation in the upstream area of the Mamasa Sub-watershed are caused by land degradation and forest conversion due to land expansion and shifting cultivation for cocoa, corn, and coffee. This study aimed to determine the amount of sediment in the Mamasa Sub-watershed in the Mamasa Regency by using the SWAT model. This research was conducted in the Mamasa Sub-watershed from June to September 2022 through several stages in the form of literature study, primary and secondary data collection. Then proceed with laboratory analysis, making a base map for the data analysis process. SWAT requires input data in land cover, soil type maps, slope maps, and climate data. The SWAT simulation was carried out in the 2012 to 2021 timeframe. The sedimentation values obtained from the SWAT model were sediment values from 0.06 to 34.073.01 tons/ha (55.34%), sediment values from 34.073.01 to 95.323.59 tons/ha (23.27%), sediment value 95.323.59 to 225.951.47 (13.51%), sediment values 225.951.47 to 442.013.09 (6.23%), and sediment values from 442.013.09 to 1,415.454.83 (1.66%). Based on this research, it can be concluded that the highest average sediment is found in the downstream area, with an area of 38,875.66 ha.

Assessment of potential soil degradation on various land uses in Keduang Watershed

Keduang watershed is a catchment area in the Wonogiri and Karanganyar districts, contributing 38.33% of sedimentation to the Gajah Mungkur Reservoir. High value population pressure is present on agricultural land in the Keduang watershed, triggering increasingly serious land degradation. Degradation in the upstream sub-watershed has a strong effect on the downstream areas. Impacts of land degradation include a decrease in land productivity and farmers’ income, as well as the occurrence of critical land. Assessment of the potential soil degradation is important as an early indicator to determine the distribution of areas that have potential soil degradation for physical, chemical, and biological soil. This research was conducted in the Keduang watershed with the land use type consisting of paddy field, wood, moor, and plantation that overlayed through land use, rainfall, slope, and soil type map. The method used in this research is guided qualification. The result showed that the Keduang watershed has a soil degradation class from low to high. The medium class dominated the potential for soil degradation in Keduang watershed with the highest potential for degradation refers to moor with a slope of >40%, rainfall of 1908 mm/year, and Inceptisols soil type.

PREDICTION OF EROSION IN HILLY AREAS OF KHILAU SUB-SUB WATERSHED USING THE RUSLE METHOD

Soil erosion poses an environmental threat to the sustainability and productivity of the land in the Khilau Sub-Sub Watershed, Bulok Sub-Watershed, Sekampung Watershed, Pesawaran Regency. The research aimed to predict soil erosion by implementing the RUSLE model on the Khilau Sub-Sub-Watershed, combined with remote sensing, GIS techniques, and field observation. This research is divided into several phases, i.e. soil data collection, getting satellite maps, obtaining rainfall data from 2011 to 2019, making the topographic maps of land slopes, land use, and soil types, whose overlay will be made, and data analysis. The results show that the areas with the highest risk of erosion were located on the hills and in the areas characterized by steep slopes of 1,395.48 (t/ha/year) with a slope level of &gt;45% (very steep), which had the highest LS factor value and were characterized as areas with strong relief, i.e. hills (with a slope level of &gt;20%). In conclusion, the erosion rate of the Khilau Sub-Sub Watershed is dominated by the erosion falling into the heavy category with an average erosion value of 405.17 (t/ha/year) by soil parameter, slope inclination, vegetation cover, and climate. The most influential factors in this study were the climate and slopes, where the lands used as mixed gardens and annuals had a high risk of erosion. So, it is of utmost importance to take soil protection and conservation measures for potential erosion protection in the area.

Investigating the Use of Street-Level Imagery and Deep Learning to Produce In-Situ Crop Type Information

The creation of crop type maps from satellite data has proven challenging and is often impeded by a lack of accurate in situ data. Street-level imagery represents a new potential source of in situ data that may aid crop type mapping, but it requires automated algorithms to recognize the features of interest. This paper aims to demonstrate a method for crop type (i.e., maize, wheat and others) recognition from street-level imagery based on a convolutional neural network using a bottom-up approach. We trained the model with a highly accurate dataset of crowdsourced labelled street-level imagery using the Picture Pile application. The classification results achieved an AUC of 0.87 for wheat, 0.85 for maize and 0.73 for others. Given that wheat and maize are two of the most common food crops grown globally, combined with an ever-increasing amount of available street-level imagery, this approach could help address the need for improved global crop type monitoring. Challenges remain in addressing the noise aspect of street-level imagery (i.e., buildings, hedgerows, automobiles, etc.) and uncertainties due to differences in the time of day and location. Such an approach could also be applied to developing other in situ data sets from street-level imagery, e.g., for land use mapping or socioeconomic indicators.

Non-Standard Map Robot Path Planning Approach Based on Ant Colony Algorithms.

Robot path planning is an important component of ensuring the robots complete work tasks effectively. Nowadays, most maps used for robot path planning obtain relevant coordinate information through sensor measurement, establish a map model based on coordinate information, and then carry out path planning for the robot, which is time-consuming and labor-intensive. To solve this problem, a method of robot path planning based on ant colony algorithms after the standardized design of non-standard map grids such as photos was studied. This method combines the robot grid map modeling with image processing, bringing in calibration objects. By converting non-standard actual environment maps into standard grid maps, this method was made suitable for robot motion path planning on non-standard maps of different types and sizes. After obtaining the planned path and pose, the robot motion path planning map under the non-standard map was obtained by combining the planned path and pose with the non-standard real environment map. The experimental results showed that this method has a high adaptability to robot non-standard map motion planning, can realize robot path planning under non-standard real environment maps, and can make the obtained robot motion path display more intuitive and convenient.