Pub Date : 2025-02-13DOI: 10.1016/j.aiig.2025.100105
Mohanad Diab , Polychronis Kolokoussis , Maria Antonia Brovelli
The use of AI technologies in remote sensing (RS) tasks has been the focus of many individuals in both the professional and academic domains. Having more accessible interfaces and tools that allow people of little or no experience to intuitively interact with RS data of multiple formats is a potential provided by this integration. However, the use of AI and AI agents to help automate RS-related tasks is still in its infancy stage, with some frameworks and interfaces built on top of well-known vision language models (VLM) such as GPT-4, segment anything model (SAM), and grounding DINO. These tools do promise and draw guidelines on the potentials and limitations of existing solutions concerning the use of said models. In this work, the state of the art AI foundation models (FM) are reviewed and used in a multi-modal manner to ingest RS imagery input and perform zero-shot object detection using natural language. The natural language input is then used to define the classes or labels the model should look for, then, both inputs are fed to the pipeline. The pipeline presented in this work makes up for the shortcomings of the general knowledge FMs by stacking pre-processing and post-processing applications on top of the FMs; these applications include tiling to produce uniform patches of the original image for faster detection, outlier rejection of redundant bounding boxes using statistical and machine learning methods. The pipeline was tested with UAV, aerial and satellite images taken over multiple areas. The accuracy for the semantic segmentation showed improvement from the original 64% to approximately 80%–99% by utilizing the pipeline and techniques proposed in this work. GitHub Repository:MohanadDiab/LangRS.
{"title":"Optimizing zero-shot text-based segmentation of remote sensing imagery using SAM and Grounding DINO","authors":"Mohanad Diab , Polychronis Kolokoussis , Maria Antonia Brovelli","doi":"10.1016/j.aiig.2025.100105","DOIUrl":"10.1016/j.aiig.2025.100105","url":null,"abstract":"<div><div>The use of AI technologies in remote sensing (RS) tasks has been the focus of many individuals in both the professional and academic domains. Having more accessible interfaces and tools that allow people of little or no experience to intuitively interact with RS data of multiple formats is a potential provided by this integration. However, the use of AI and AI agents to help automate RS-related tasks is still in its infancy stage, with some frameworks and interfaces built on top of well-known vision language models (VLM) such as GPT-4, segment anything model (SAM), and grounding DINO. These tools do promise and draw guidelines on the potentials and limitations of existing solutions concerning the use of said models. In this work, the state of the art AI foundation models (FM) are reviewed and used in a multi-modal manner to ingest RS imagery input and perform zero-shot object detection using natural language. The natural language input is then used to define the classes or labels the model should look for, then, both inputs are fed to the pipeline. The pipeline presented in this work makes up for the shortcomings of the general knowledge FMs by stacking pre-processing and post-processing applications on top of the FMs; these applications include tiling to produce uniform patches of the original image for faster detection, outlier rejection of redundant bounding boxes using statistical and machine learning methods. The pipeline was tested with UAV, aerial and satellite images taken over multiple areas. The accuracy for the semantic segmentation showed improvement from the original 64% to approximately 80%–99% by utilizing the pipeline and techniques proposed in this work. <strong>GitHub Repository:</strong> <span><span>MohanadDiab/LangRS</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"6 1","pages":"Article 100105"},"PeriodicalIF":0.0,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143429918","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-27DOI: 10.1016/j.aiig.2025.100106
Kaoutar Clero , Said Ed-Diny , Mohammed Achalhi , Mouhamed Cherkaoui , Imad El Harraki , Sanaa El Fkihi , Intissar Benzakour , Tarik Soror , Said Rziki , Hamd Ait Abdelali , Hicham Tagemouati , François Bourzeix
Rockfalls are among the frequent hazards in underground mines worldwide, requiring effective methods for detecting unstable rock blocks to ensure miners' and equipment's safety. This study proposes a novel approach for identifying potential rockfall zones using infrared thermal imaging and image segmentation techniques. Infrared images of rock blocks were captured at the Draa Sfar deep underground mine in Morocco using the FLUKE TI401 PRO thermal camera. Two segmentation methods were applied to locate the potential unstable areas: the classical thresholding and the K-means clustering model. The results show that while thresholding allows a binary distinction between stable and unstable areas, K-means clustering is more accurate, especially when using multiple clusters to show different risk levels. The close match between the clustering masks of unstable blocks and their corresponding visible light images further validated this. The findings confirm that thermal image segmentation can serve as an alternative method for predicting rockfalls and monitoring geotechnical issues in underground mines. Underground operators worldwide can apply this approach to monitor rock mass stability. However, further research is recommended to enhance these results, particularly through deep learning-based segmentation and object detection models.
{"title":"Loosening rocks detection at Draa Sfar deep underground mine in Morocco using infrared thermal imaging and image segmentation models","authors":"Kaoutar Clero , Said Ed-Diny , Mohammed Achalhi , Mouhamed Cherkaoui , Imad El Harraki , Sanaa El Fkihi , Intissar Benzakour , Tarik Soror , Said Rziki , Hamd Ait Abdelali , Hicham Tagemouati , François Bourzeix","doi":"10.1016/j.aiig.2025.100106","DOIUrl":"10.1016/j.aiig.2025.100106","url":null,"abstract":"<div><div>Rockfalls are among the frequent hazards in underground mines worldwide, requiring effective methods for detecting unstable rock blocks to ensure miners' and equipment's safety. This study proposes a novel approach for identifying potential rockfall zones using infrared thermal imaging and image segmentation techniques. Infrared images of rock blocks were captured at the Draa Sfar deep underground mine in Morocco using the FLUKE TI401 PRO thermal camera. Two segmentation methods were applied to locate the potential unstable areas: the classical thresholding and the K-means clustering model. The results show that while thresholding allows a binary distinction between stable and unstable areas, K-means clustering is more accurate, especially when using multiple clusters to show different risk levels. The close match between the clustering masks of unstable blocks and their corresponding visible light images further validated this. The findings confirm that thermal image segmentation can serve as an alternative method for predicting rockfalls and monitoring geotechnical issues in underground mines. Underground operators worldwide can apply this approach to monitor rock mass stability. However, further research is recommended to enhance these results, particularly through deep learning-based segmentation and object detection models.</div></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"6 1","pages":"Article 100106"},"PeriodicalIF":0.0,"publicationDate":"2025-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143143632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-04DOI: 10.1016/j.aiig.2024.100104
Qiansheng Wei , Zishuai Li , Haonan Feng , Yueying Jiang , Yang Yang , Zhiguo Wang
Seismic time-frequency (TF) transforms are essential tools in reservoir interpretation and signal processing, particularly for characterizing frequency variations in non-stationary seismic data. Recently, sparse TF transforms, which leverage sparse coding (SC), have gained significant attention in the geosciences due to their ability to achieve high TF resolution. However, the iterative approaches typically employed in sparse TF transforms are computationally intensive, making them impractical for real seismic data analysis. To address this issue, we propose an interpretable convolutional sparse coding (CSC) network to achieve high TF resolution. The proposed model is generated based on the traditional short-time Fourier transform (STFT) transform and a modified UNet, named ULISTANet. In this design, we replace the conventional convolutional layers of the UNet with learnable iterative shrinkage thresholding algorithm (LISTA) blocks, a specialized form of CSC. The LISTA block, which evolves from the traditional iterative shrinkage thresholding algorithm (ISTA), is optimized for extracting sparse features more effectively. Furthermore, we create a synthetic dataset featuring complex frequency-modulated signals to train ULISTANet. Finally, the proposed method's performance is subsequently validated using both synthetic and field data, demonstrating its potential for enhanced seismic data analysis.
{"title":"Convolutional sparse coding network for sparse seismic time-frequency representation","authors":"Qiansheng Wei , Zishuai Li , Haonan Feng , Yueying Jiang , Yang Yang , Zhiguo Wang","doi":"10.1016/j.aiig.2024.100104","DOIUrl":"10.1016/j.aiig.2024.100104","url":null,"abstract":"<div><div>Seismic time-frequency (TF) transforms are essential tools in reservoir interpretation and signal processing, particularly for characterizing frequency variations in non-stationary seismic data. Recently, sparse TF transforms, which leverage sparse coding (SC), have gained significant attention in the geosciences due to their ability to achieve high TF resolution. However, the iterative approaches typically employed in sparse TF transforms are computationally intensive, making them impractical for real seismic data analysis. To address this issue, we propose an interpretable convolutional sparse coding (CSC) network to achieve high TF resolution. The proposed model is generated based on the traditional short-time Fourier transform (STFT) transform and a modified UNet, named ULISTANet. In this design, we replace the conventional convolutional layers of the UNet with learnable iterative shrinkage thresholding algorithm (LISTA) blocks, a specialized form of CSC. The LISTA block, which evolves from the traditional iterative shrinkage thresholding algorithm (ISTA), is optimized for extracting sparse features more effectively. Furthermore, we create a synthetic dataset featuring complex frequency-modulated signals to train ULISTANet. Finally, the proposed method's performance is subsequently validated using both synthetic and field data, demonstrating its potential for enhanced seismic data analysis.</div></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"6 1","pages":"Article 100104"},"PeriodicalIF":0.0,"publicationDate":"2024-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142661128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-16DOI: 10.1016/j.aiig.2024.100095
Cai Li, Fei Ma, Yuxiu Wang, Delong Zhang
The connectivity of sandbodies is a key constraint to the exploration effectiveness of Bohai A Oilfield. Conventional connectivity studies often use methods such as seismic attribute fusion, while the development of contiguous composite sandbodies in this area makes it challenging to characterize connectivity changes with conventional seismic attributes. Aiming at the above problem in the Bohai A Oilfield, this study proposes a big data analysis method based on the Deep Forest algorithm to predict the sandbody connectivity. Firstly, by compiling the abundant exploration and development sandbodies data in the study area, typical sandbodies with reliable connectivity were selected. Then, sensitive seismic attribute were extracted to obtain training samples. Finally, based on the Deep Forest algorithm, mapping model between attribute combinations and sandbody connectivity was established through machine learning. This method achieves the first quantitative determination of the connectivity for continuous composite sandbodies in the Bohai Oilfield. Compared with conventional connectivity discrimination methods such as high-resolution processing and seismic attribute analysis, this method can combine the sandbody characteristics of the study area in the process of machine learning, and jointly judge connectivity by combining multiple seismic attributes. The study results show that this method has high accuracy and timeliness in predicting connectivity for continuous composite sandbodies. Applied to the Bohai A Oilfield, it successfully identified multiple sandbody connectivity relationships and provided strong support for the subsequent exploration potential assessment and well placement optimization. This method also provides a new idea and method for studying sandbody connectivity under similar complex geological conditions.
砂体连通性是渤海 A 油田勘探有效性的关键制约因素。传统的连通性研究通常采用地震属性融合等方法，而该地区连片复合砂体的发育使得用常规地震属性表征连通性变化具有挑战性。针对渤海A油田的上述问题，本研究提出了一种基于深林算法的大数据分析方法来预测砂体连通性。首先，通过整理研究区丰富的勘探开发沙体数据，筛选出具有可靠连通性的典型沙体。然后，提取敏感地震属性，获得训练样本。最后，基于深林算法，通过机器学习建立属性组合与沙体连通性之间的映射模型。该方法首次实现了对渤海油田连续复合砂体连通性的定量判定。与传统的高分辨率处理、地震属性分析等连通性判别方法相比，该方法在机器学习过程中能够结合研究区的沙体特征，综合多种地震属性共同判断连通性。研究结果表明，该方法在预测连续复合砂体连通性方面具有较高的准确性和时效性。应用于渤海A油田，成功识别了多个砂体的连通性关系，为后续的勘探潜力评估和井位优化提供了有力支持。该方法也为研究类似复杂地质条件下的砂体连通性提供了新的思路和方法。
{"title":"Research on the prediction method for fluvial-phase sandbody connectivity based on big data analysis--a case study of Bohai a oilfield","authors":"Cai Li, Fei Ma, Yuxiu Wang, Delong Zhang","doi":"10.1016/j.aiig.2024.100095","DOIUrl":"10.1016/j.aiig.2024.100095","url":null,"abstract":"<div><div>The connectivity of sandbodies is a key constraint to the exploration effectiveness of Bohai A Oilfield. Conventional connectivity studies often use methods such as seismic attribute fusion, while the development of contiguous composite sandbodies in this area makes it challenging to characterize connectivity changes with conventional seismic attributes. Aiming at the above problem in the Bohai A Oilfield, this study proposes a big data analysis method based on the Deep Forest algorithm to predict the sandbody connectivity. Firstly, by compiling the abundant exploration and development sandbodies data in the study area, typical sandbodies with reliable connectivity were selected. Then, sensitive seismic attribute were extracted to obtain training samples. Finally, based on the Deep Forest algorithm, mapping model between attribute combinations and sandbody connectivity was established through machine learning. This method achieves the first quantitative determination of the connectivity for continuous composite sandbodies in the Bohai Oilfield. Compared with conventional connectivity discrimination methods such as high-resolution processing and seismic attribute analysis, this method can combine the sandbody characteristics of the study area in the process of machine learning, and jointly judge connectivity by combining multiple seismic attributes. The study results show that this method has high accuracy and timeliness in predicting connectivity for continuous composite sandbodies. Applied to the Bohai A Oilfield, it successfully identified multiple sandbody connectivity relationships and provided strong support for the subsequent exploration potential assessment and well placement optimization. This method also provides a new idea and method for studying sandbody connectivity under similar complex geological conditions.</div></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"5 ","pages":"Article 100095"},"PeriodicalIF":0.0,"publicationDate":"2024-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142532292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pore size analysis plays a pivotal role in unraveling reservoir behavior and its intricate relationship with confined fluids. Traditional methods for predicting pore size distribution (PSD), relying on drilling cores or thin sections, face limitations associated with depth specificity. In this study, we introduce an innovative framework that leverages nuclear magnetic resonance (NMR) log data, encompassing clay-bound water (CBW), bound volume irreducible (BVI), and free fluid volume (FFV), to determine three PSDs (micropores, mesopores, and macropores). Moreover, we establish a robust pore size classification (PSC) system utilizing ternary plots, derived from the PSDs.
Within the three studied wells, NMR log data is exclusive to one well (well-A), while conventional well logs are accessible for all three wells (well-A, well-B, and well-C). This distinction enables PSD predictions for the remaining two wells (B and C). To prognosticate NMR outputs (CBW, BVI, FFV) for these wells, a two-step deep learning (DL) algorithm is implemented. Initially, three feature selection algorithms (f-classif, f-regression, and mutual-info-regression) identify the conventional well logs most correlated to NMR outputs in well-A. The three feature selection algorithms utilize statistical computations. These algorithms are utilized to systematically identify and optimize pertinent input features, thereby augmenting model interpretability and predictive efficacy within intricate data-driven endeavors. So, all three feature selection algorithms introduced the number of 4 logs as the most optimal number of inputs to the DL algorithm with different combinations of logs for each of the three desired outputs. Subsequently, the CUDA Deep Neural Network Long Short-Term Memory algorithm(CUDNNLSTM), belonging to the category of DL algorithms and harnessing the computational power of GPUs, is employed for the prediction of CBW, BVI, and FFV logs. This prediction leverages the optimal logs identified in the preceding step. Estimation of NMR outputs was done first in well-A (80% of data as training and 20% as testing). The correlation coefficient (CC) between the actual and estimated data for the three outputs CBW, BVI and FFV are 95%, 94%, and 97%, respectively, as well as root mean square error (RMSE) was obtained 0.0081, 0.098, and 0.0089, respectively. To assess the effectiveness of the proposed algorithm, we compared it with two traditional methods for log estimation: multiple regression and multi-resolution graph-based clustering methods. The results demonstrate the superior accuracy of our algorithm in comparison to these conventional approaches. This DL-driven approach facilitates PSD prediction grounded in fluid saturation for wells B and C.
Ternary plots are then employed for PSCs. Seven distinct PSCs within well-A employing actual NMR logs (CBW, BVI, FFV), in conjunction with an equivalent count within wells B and C utilizing three predicted
孔隙度分析在揭示储层行为及其与封闭流体的复杂关系方面起着举足轻重的作用。传统的孔径分布（PSD）预测方法依赖于钻井岩心或薄切片，面临着深度特异性的限制。在这项研究中，我们引入了一个创新框架，利用核磁共振（NMR）测井数据（包括粘土结合水（CBW）、不可还原结合体积（BVI）和自由流体体积（FFV））来确定三种孔径分布（微孔、中孔和大孔）。此外，我们还利用从 PSDs 得出的三元图建立了一个强大的孔径分类 (PSC) 系统。在所研究的三口井中，核磁共振测井数据仅适用于一口井（A 井），而常规测井数据则适用于所有三口井（A 井、B 井和 C 井）。这种区别使我们能够对其余两口井（B 井和 C 井）进行 PSD 预测。为了预测这些油井的 NMR 输出（CBW、BVI、FFV），采用了两步深度学习（DL）算法。首先，三种特征选择算法（f-classif、f-regression 和 mutual-info-regression）确定与 A 井 NMR 输出最相关的常规测井曲线。这三种特征选择算法利用统计计算。这些算法用于系统地识别和优化相关输入特征，从而在复杂的数据驱动工作中提高模型的可解释性和预测效力。因此，所有三种特征选择算法都将 4 个日志的数量作为 DL 算法的最佳输入数量，并为三种所需的输出分别引入不同的日志组合。随后，CUDA 深度神经网络长短期记忆算法（CUDNNLSTM）被用于预测 CBW、BVI 和 FFV 日志，该算法属于 DL 算法范畴，利用了 GPU 的计算能力。该预测利用了前一步中确定的最佳日志。首先在 A 井（80% 的数据作为训练数据，20% 作为测试数据）中对 NMR 输出进行估计。CBW、BVI 和 FFV 三项输出的实际数据与估计数据之间的相关系数（CC）分别为 95%、94% 和 97%，均方根误差（RMSE）分别为 0.0081、0.098 和 0.0089。为了评估所提出算法的有效性，我们将其与两种传统的对数估计方法进行了比较：多元回归和基于多分辨率图的聚类方法。结果表明，与这些传统方法相比，我们的算法具有更高的准确性。这种以 DL 为驱动的方法有助于根据 B 井和 C 井的流体饱和度预测 PSD。利用实际 NMR 测井（CBW、BVI、FFV）对 A 井中的七个不同 PSC 进行了和谐分类，同时利用三个预测测井对 B 井和 C 井中的等量 PSC 进行了和谐分类，从而确定了七个不同的孔径分类面 (PSCF)。由此产生的 PSCF 为生成精确、详细的储层三维模型提供了宝贵的见解。
{"title":"Pore size classification and prediction based on distribution of reservoir fluid volumes utilizing well logs and deep learning algorithm in a complex lithology","authors":"Hassan Bagheri , Reza Mohebian , Ali Moradzadeh , Behnia Azizzadeh Mehmandost Olya","doi":"10.1016/j.aiig.2024.100094","DOIUrl":"10.1016/j.aiig.2024.100094","url":null,"abstract":"<div><div>Pore size analysis plays a pivotal role in unraveling reservoir behavior and its intricate relationship with confined fluids. Traditional methods for predicting pore size distribution (PSD), relying on drilling cores or thin sections, face limitations associated with depth specificity. In this study, we introduce an innovative framework that leverages nuclear magnetic resonance (NMR) log data, encompassing clay-bound water (CBW), bound volume irreducible (BVI), and free fluid volume (FFV), to determine three PSDs (micropores, mesopores, and macropores). Moreover, we establish a robust pore size classification (PSC) system utilizing ternary plots, derived from the PSDs.</div><div>Within the three studied wells, NMR log data is exclusive to one well (well-A), while conventional well logs are accessible for all three wells (well-A, well-B, and well-C). This distinction enables PSD predictions for the remaining two wells (B and C). To prognosticate NMR outputs (CBW, BVI, FFV) for these wells, a two-step deep learning (DL) algorithm is implemented. Initially, three feature selection algorithms (f-classif, f-regression, and mutual-info-regression) identify the conventional well logs most correlated to NMR outputs in well-A. The three feature selection algorithms utilize statistical computations. These algorithms are utilized to systematically identify and optimize pertinent input features, thereby augmenting model interpretability and predictive efficacy within intricate data-driven endeavors. So, all three feature selection algorithms introduced the number of 4 logs as the most optimal number of inputs to the DL algorithm with different combinations of logs for each of the three desired outputs. Subsequently, the CUDA Deep Neural Network Long Short-Term Memory algorithm(CUDNNLSTM), belonging to the category of DL algorithms and harnessing the computational power of GPUs, is employed for the prediction of CBW, BVI, and FFV logs. This prediction leverages the optimal logs identified in the preceding step. Estimation of NMR outputs was done first in well-A (80% of data as training and 20% as testing). The correlation coefficient (CC) between the actual and estimated data for the three outputs CBW, BVI and FFV are 95%, 94%, and 97%, respectively, as well as root mean square error (RMSE) was obtained 0.0081, 0.098, and 0.0089, respectively. To assess the effectiveness of the proposed algorithm, we compared it with two traditional methods for log estimation: multiple regression and multi-resolution graph-based clustering methods. The results demonstrate the superior accuracy of our algorithm in comparison to these conventional approaches. This DL-driven approach facilitates PSD prediction grounded in fluid saturation for wells B and C.</div><div>Ternary plots are then employed for PSCs. Seven distinct PSCs within well-A employing actual NMR logs (CBW, BVI, FFV), in conjunction with an equivalent count within wells B and C utilizing three predicted","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"5 ","pages":"Article 100094"},"PeriodicalIF":0.0,"publicationDate":"2024-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142437701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-05DOI: 10.1016/j.aiig.2024.100093
Guruh Samodra , Ngadisih , Ferman Setia Nugroho
Machine learning (ML) algorithms are frequently used in landslide susceptibility modeling. Different data handling strategies may generate variations in landslide susceptibility modeling, even when using the same ML algorithm. This research aims to compare the combinations of inventory data handling, cross validation (CV), and hyperparameter tuning strategies to generate landslide susceptibility maps. The results are expected to provide a general strategy for landslide susceptibility modeling using ML techniques. The authors employed eight landslide inventory data handling scenarios to convert a landslide polygon into a landslide point, i.e., the landslide point is located on the toe (minimum height), on the scarp (maximum height), at the center of the landslide, randomly inside the polygon (1 point), randomly inside the polygon (3 points), randomly inside the polygon (5 points), randomly inside the polygon (10 points), and 15 m grid sampling. Random forest models using CV–nonspatial hyperparameter tuning, spatial CV–spatial hyperparameter tuning, and spatial CV–forward feature selection–no hyperparameter tuning were applied for each data handling strategy. The combination generated 24 random forest ML workflows, which are applied using a complete inventory of 743 landslides triggered by Tropical Cyclone Cempaka (2017) in Pacitan Regency, Indonesia, and 11 landslide controlling factors. The results show that grid sampling with spatial CV and spatial hyperparameter tuning is favorable because the strategy can minimize overfitting, generate a relatively high-performance predictive model, and reduce the appearance of susceptibility artifacts in the landslide area. Careful data inventory handling, CV, and hyperparameter tuning strategies should be considered in landslide susceptibility modeling to increase the applicability of landslide susceptibility maps in practical application.
机器学习（ML）算法经常用于滑坡易感性建模。即使使用相同的 ML 算法，不同的数据处理策略也可能导致滑坡易感性建模的差异。本研究旨在比较库存数据处理、交叉验证（CV）和超参数调整策略的组合，以生成滑坡易感性图。研究结果有望为使用 ML 技术进行滑坡易感性建模提供通用策略。作者采用了八种滑坡清单数据处理方案，将滑坡多边形转换为滑坡点，即滑坡点位于坡脚（最小高度）、坡面（最大高度）、滑坡中心、多边形内随机（1 点）、多边形内随机（3 点）、多边形内随机（5 点）、多边形内随机（10 点）和 15 米网格采样。每种数据处理策略都采用了 CV-非空间超参数调整、空间 CV-空间超参数调整和空间 CV-前向特征选择-无超参数调整的随机森林模型。组合生成了 24 个随机森林 ML 工作流，并将其应用于印尼帕契坦地区热带气旋 "肯帕卡"（2017 年）引发的 743 次滑坡的完整清单和 11 个滑坡控制因素。结果表明，网格采样加上空间 CV 和空间超参数调整是有利的，因为该策略可以最大限度地减少过拟合，生成性能相对较高的预测模型，并减少滑坡区域易感性假象的出现。在滑坡易感性建模中应考虑谨慎的数据清单处理、CV 和超参数调整策略，以提高滑坡易感性图在实际应用中的适用性。
{"title":"Benchmarking data handling strategies for landslide susceptibility modeling using random forest workflows","authors":"Guruh Samodra , Ngadisih , Ferman Setia Nugroho","doi":"10.1016/j.aiig.2024.100093","DOIUrl":"10.1016/j.aiig.2024.100093","url":null,"abstract":"<div><div>Machine learning (ML) algorithms are frequently used in landslide susceptibility modeling. Different data handling strategies may generate variations in landslide susceptibility modeling, even when using the same ML algorithm. This research aims to compare the combinations of inventory data handling, cross validation (CV), and hyperparameter tuning strategies to generate landslide susceptibility maps. The results are expected to provide a general strategy for landslide susceptibility modeling using ML techniques. The authors employed eight landslide inventory data handling scenarios to convert a landslide polygon into a landslide point, i.e., the landslide point is located on the toe (minimum height), on the scarp (maximum height), at the center of the landslide, randomly inside the polygon (1 point), randomly inside the polygon (3 points), randomly inside the polygon (5 points), randomly inside the polygon (10 points), and 15 m grid sampling. Random forest models using CV–nonspatial hyperparameter tuning, spatial CV–spatial hyperparameter tuning, and spatial CV–forward feature selection–no hyperparameter tuning were applied for each data handling strategy. The combination generated 24 random forest ML workflows, which are applied using a complete inventory of 743 landslides triggered by Tropical Cyclone Cempaka (2017) in Pacitan Regency, Indonesia, and 11 landslide controlling factors. The results show that grid sampling with spatial CV and spatial hyperparameter tuning is favorable because the strategy can minimize overfitting, generate a relatively high-performance predictive model, and reduce the appearance of susceptibility artifacts in the landslide area. Careful data inventory handling, CV, and hyperparameter tuning strategies should be considered in landslide susceptibility modeling to increase the applicability of landslide susceptibility maps in practical application.</div></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"5 ","pages":"Article 100093"},"PeriodicalIF":0.0,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142532855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-22DOI: 10.1016/j.aiig.2024.100092
Haiying Fu , Shuai Wang , Guicheng He , Zhonghua Zhu , Qing Yu , Dexin Ding
Porosity, tortuosity, specific surface area (SSA), and permeability are four key parameters of reactive transport modeling in sandstone, which are important for understanding solute transport and geochemical reaction processes in sandstone aquifers. These four parameters reflect the characteristics of pore structure of sandstone from different perspectives, and the traditional empirical formulas cannot make accurate predictions of them due to their complexity and heterogeneity. In this paper, eleven types of sandstone CT images were firstly segmented into numerous subsample images, the porosity, tortuosity, SSA, and permeability of the subsamples were calculated, and the dataset was established. The 3D convolutional neural network (CNN) models were subsequently established and trained to predict the key reactive transport parameters based on subsample CT images of sandstones. The results demonstrated that the 3D CNN model with multiple outputs exhibited excellent prediction ability for the four parameters compared to the traditional empirical formulas. In particular, for the prediction of tortuosity and permeability, the 3D CNN model with multiple outputs even showed slightly better prediction ability than its single-output variant model. Additionally, it demonstrated good generalization performance on sandstone CT images not included in the training dataset. The study showed that the 3D CNN model with multiple outputs has the advantages of simplifying operation and saving computational resources, which has the prospect of popularization and application.
{"title":"A 3D convolutional neural network model with multiple outputs for simultaneously estimating the reactive transport parameters of sandstone from its CT images","authors":"Haiying Fu , Shuai Wang , Guicheng He , Zhonghua Zhu , Qing Yu , Dexin Ding","doi":"10.1016/j.aiig.2024.100092","DOIUrl":"10.1016/j.aiig.2024.100092","url":null,"abstract":"<div><div>Porosity, tortuosity, specific surface area (SSA), and permeability are four key parameters of reactive transport modeling in sandstone, which are important for understanding solute transport and geochemical reaction processes in sandstone aquifers. These four parameters reflect the characteristics of pore structure of sandstone from different perspectives, and the traditional empirical formulas cannot make accurate predictions of them due to their complexity and heterogeneity. In this paper, eleven types of sandstone CT images were firstly segmented into numerous subsample images, the porosity, tortuosity, SSA, and permeability of the subsamples were calculated, and the dataset was established. The 3D convolutional neural network (CNN) models were subsequently established and trained to predict the key reactive transport parameters based on subsample CT images of sandstones. The results demonstrated that the 3D CNN model with multiple outputs exhibited excellent prediction ability for the four parameters compared to the traditional empirical formulas. In particular, for the prediction of tortuosity and permeability, the 3D CNN model with multiple outputs even showed slightly better prediction ability than its single-output variant model. Additionally, it demonstrated good generalization performance on sandstone CT images not included in the training dataset. The study showed that the 3D CNN model with multiple outputs has the advantages of simplifying operation and saving computational resources, which has the prospect of popularization and application.</div></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"5 ","pages":"Article 100092"},"PeriodicalIF":0.0,"publicationDate":"2024-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142318693","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-20DOI: 10.1016/j.aiig.2024.100091
Binsen Xu , Zhou Feng , Jun Zhou , Rongbo Shao , Hongliang Wu , Peng Liu , Han Tian , Weizhong Li , Lizhi Xiao
Machine learning has been widely applied in well logging formation evaluation studies. However, several challenges negatively impacted the generalization capabilities of machine learning models in practical implementations, such as the mismatch of data domain between training and testing datasets, imbalances among sample categories, and inadequate representation of data model. These issues have led to substantial insufficient identification for reservoir and significant deviations in subsequent evaluations. To improve the transferability of machine learning models within limited sample sets, this study proposes a weight transfer learning framework based on the similarity of the labels. The similarity weighting method includes both hard weights and soft weights. By evaluating the similarity between test and training sets of logging data, the similarity results are used to estimate the weights of training samples, thereby optimizing the model learning process. We develop a double experts' network and a bidirectional gated neural network based on hierarchical attention and multi-head attention (BiGRU-MHSA) for well logs reconstruction and lithofacies classification tasks. Oil field data results for the shale strata in the Gulong area of the Songliao Basin of China indicate that the double experts’ network model performs well in curve reconstruction tasks. However, it may not be effective in lithofacies classification tasks, while BiGRU-MHSA performs well in that area. In the study of constructing large-scale well logging processing and formation interpretation models, it is maybe more beneficial by employing different expert models for combined evaluations. In addition, although the improvement is limited, hard or soft weighting methods is better than unweighted (i.e., average-weighted) in significantly different adjacent wells. The code and data are open and available for subsequent studies on other lithofacies layers.
{"title":"Transfer learning for well logging formation evaluation using similarity weights","authors":"Binsen Xu , Zhou Feng , Jun Zhou , Rongbo Shao , Hongliang Wu , Peng Liu , Han Tian , Weizhong Li , Lizhi Xiao","doi":"10.1016/j.aiig.2024.100091","DOIUrl":"10.1016/j.aiig.2024.100091","url":null,"abstract":"<div><div>Machine learning has been widely applied in well logging formation evaluation studies. However, several challenges negatively impacted the generalization capabilities of machine learning models in practical implementations, such as the mismatch of data domain between training and testing datasets, imbalances among sample categories, and inadequate representation of data model. These issues have led to substantial insufficient identification for reservoir and significant deviations in subsequent evaluations. To improve the transferability of machine learning models within limited sample sets, this study proposes a weight transfer learning framework based on the similarity of the labels. The similarity weighting method includes both hard weights and soft weights. By evaluating the similarity between test and training sets of logging data, the similarity results are used to estimate the weights of training samples, thereby optimizing the model learning process. We develop a double experts' network and a bidirectional gated neural network based on hierarchical attention and multi-head attention (BiGRU-MHSA) for well logs reconstruction and lithofacies classification tasks. Oil field data results for the shale strata in the Gulong area of the Songliao Basin of China indicate that the double experts’ network model performs well in curve reconstruction tasks. However, it may not be effective in lithofacies classification tasks, while BiGRU-MHSA performs well in that area. In the study of constructing large-scale well logging processing and formation interpretation models, it is maybe more beneficial by employing different expert models for combined evaluations. In addition, although the improvement is limited, hard or soft weighting methods is better than unweighted (i.e., average-weighted) in significantly different adjacent wells. The code and data are open and available for subsequent studies on other lithofacies layers.</div></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"5 ","pages":"Article 100091"},"PeriodicalIF":0.0,"publicationDate":"2024-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurately and efficiently predicting the permeability of porous media is essential for addressing a wide range of hydrogeological issues. However, the complexity of porous media often limits the effectiveness of individual prediction methods. This study introduces a novel Particle Swarm Optimization-based Permeability Integrated Prediction model (PSO-PIP), which incorporates a particle swarm optimization algorithm enhanced with dynamic clustering and adaptive parameter tuning (KGPSO). The model integrates multi-source data from the Lattice Boltzmann Method (LBM), Pore Network Modeling (PNM), and Finite Difference Method (FDM). By assigning optimal weight coefficients to the outputs of these methods, the model minimizes deviations from actual values and enhances permeability prediction performance. Initially, the computational performances of the LBM, PNM, and FDM are comparatively analyzed on datasets consisting of sphere packings and real rock samples. It is observed that these methods exhibit computational biases in certain permeability ranges. The PSO-PIP model is proposed to combine the strengths of each computational approach and mitigate their limitations. The PSO-PIP model consistently produces predictions that are highly congruent with actual permeability values across all prediction intervals, significantly enhancing prediction accuracy. The outcomes of this study provide a new tool and perspective for the comprehensive, rapid, and accurate prediction of permeability in porous media.
{"title":"Enhanced permeability prediction in porous media using particle swarm optimization with multi-source integration","authors":"Zhiping Chen , Jia Zhang , Daren Zhang , Xiaolin Chang , Wei Zhou","doi":"10.1016/j.aiig.2024.100090","DOIUrl":"10.1016/j.aiig.2024.100090","url":null,"abstract":"<div><p>Accurately and efficiently predicting the permeability of porous media is essential for addressing a wide range of hydrogeological issues. However, the complexity of porous media often limits the effectiveness of individual prediction methods. This study introduces a novel Particle Swarm Optimization-based Permeability Integrated Prediction model (PSO-PIP), which incorporates a particle swarm optimization algorithm enhanced with dynamic clustering and adaptive parameter tuning (KGPSO). The model integrates multi-source data from the Lattice Boltzmann Method (LBM), Pore Network Modeling (PNM), and Finite Difference Method (FDM). By assigning optimal weight coefficients to the outputs of these methods, the model minimizes deviations from actual values and enhances permeability prediction performance. Initially, the computational performances of the LBM, PNM, and FDM are comparatively analyzed on datasets consisting of sphere packings and real rock samples. It is observed that these methods exhibit computational biases in certain permeability ranges. The PSO-PIP model is proposed to combine the strengths of each computational approach and mitigate their limitations. The PSO-PIP model consistently produces predictions that are highly congruent with actual permeability values across all prediction intervals, significantly enhancing prediction accuracy. The outcomes of this study provide a new tool and perspective for the comprehensive, rapid, and accurate prediction of permeability in porous media.</p></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"5 ","pages":"Article 100090"},"PeriodicalIF":0.0,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666544124000315/pdfft?md5=b1d09a2cb0aeba96843adc16601b4089&pid=1-s2.0-S2666544124000315-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142271080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-24DOI: 10.1016/j.aiig.2024.100087
Bas Peters , Eldad Haber , Keegan Lensink
The large spatial/temporal/frequency scale of geoscience and remote-sensing datasets causes memory issues when using convolutional neural networks for (sub-) surface data segmentation. Recently developed fully reversible or fully invertible networks can mostly avoid memory limitations by recomputing the states during the backward pass through the network. This results in a low and fixed memory requirement for storing network states, as opposed to the typical linear memory growth with network depth. This work focuses on a fully invertible network based on the telegraph equation. While reversibility saves the major amount of memory used in deep networks by the data, the convolutional kernels can take up most memory if fully invertible networks contain multiple invertible pooling/coarsening layers. We address the explosion of the number of convolutional kernels by combining fully invertible networks with layers that contain the convolutional kernels in a compressed form directly. A second challenge is that invertible networks output a tensor the same size as its input. This property prevents the straightforward application of invertible networks to applications that map between different input–output dimensions, need to map to outputs with more channels than present in the input data, or desire outputs that decrease/increase the resolution compared to the input data. However, we show that by employing invertible networks in a non-standard fashion, we can still use them for these tasks. Examples in hyperspectral land-use classification, airborne geophysical surveying, and seismic imaging illustrate that we can input large data volumes in one chunk and do not need to work on small patches, use dimensionality reduction, or employ methods that classify a patch to a single central pixel.
{"title":"Fully invertible hyperbolic neural networks for segmenting large-scale surface and sub-surface data","authors":"Bas Peters , Eldad Haber , Keegan Lensink","doi":"10.1016/j.aiig.2024.100087","DOIUrl":"10.1016/j.aiig.2024.100087","url":null,"abstract":"<div><p>The large spatial/temporal/frequency scale of geoscience and remote-sensing datasets causes memory issues when using convolutional neural networks for (sub-) surface data segmentation. Recently developed fully reversible or fully invertible networks can mostly avoid memory limitations by recomputing the states during the backward pass through the network. This results in a low and fixed memory requirement for storing network states, as opposed to the typical linear memory growth with network depth. This work focuses on a fully invertible network based on the telegraph equation. While reversibility saves the major amount of memory used in deep networks by the data, the convolutional kernels can take up most memory if fully invertible networks contain multiple invertible pooling/coarsening layers. We address the explosion of the number of convolutional kernels by combining fully invertible networks with layers that contain the convolutional kernels in a compressed form directly. A second challenge is that invertible networks output a tensor the same size as its input. This property prevents the straightforward application of invertible networks to applications that map between different input–output dimensions, need to map to outputs with more channels than present in the input data, or desire outputs that decrease/increase the resolution compared to the input data. However, we show that by employing invertible networks in a non-standard fashion, we can still use them for these tasks. Examples in hyperspectral land-use classification, airborne geophysical surveying, and seismic imaging illustrate that we can input large data volumes in one chunk and do not need to work on small patches, use dimensionality reduction, or employ methods that classify a patch to a single central pixel.</p></div>","PeriodicalId":100124,"journal":{"name":"Artificial Intelligence in Geosciences","volume":"5 ","pages":"Article 100087"},"PeriodicalIF":0.0,"publicationDate":"2024-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666544124000285/pdfft?md5=aefb3645cc92ad5ad25d7d3f97a32057&pid=1-s2.0-S2666544124000285-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142087583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号