Deep learning based approach for the instance segmentation of clayey soil desiccation cracks

Han, Xiao-Le; Jiang, Ning-Jun; Yang, Yufei; Choi, Jong Sun; Singh, Devandra N.; Beta, Priyanka; Du, Yukou; Wang, Yijie

doi:10.1016/j.compgeo.2022.104733

Cited by 22 publications

(4 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Qualitative and quantitative studies by statistical means are necessary in order to investigate the effectiveness of the instance segmentation model and the algorithms for measuring DBH proposed in this study. In our analysis, we resorted to three main evaluation metrics to measure the accuracy of the detection and segmentation tasks, namely, AP 50:5:95 , AP 50 [45], and F 1 score [46]. The AP 50 we adopt is a widely used metric in the field of target detection, which represents the Average Precision at an Intersection over Union (IoU) of 0.5, as shown in Equations ( 8) and (9), where: TP represents True Positives, FP represents False Positives, and FN represents False Negatives.…”

Section: Data Evaluationmentioning

confidence: 99%

An Advanced Software Platform and Algorithmic Framework for Mobile DBH Data Acquisition

Zhang,

Liang,

Tong

et al. 2023

Forests

View full text Add to dashboard Cite

Rapid and precise tree Diameter at Breast Height (DBH) measurement is pivotal in forest inventories. While the recent advancements in LiDAR and Structure from Motion (SFM) technologies have paved the way for automated DBH measurements, the significant equipment costs and the complexity of operational procedures continue to constrain the ubiquitous adoption of these technologies for real-time DBH assessments. In this research, we introduce KAN-Forest, a real-time DBH measurement and key point localization algorithm utilizing RGB-D (Red, Green, Blue-Depth) imaging technology. Firstly, we improved the YOLOv5-seg segmentation module with a Channel and Spatial Attention (CBAM) module, augmenting its efficiency in extracting the tree’s edge features in intricate forest scenarios. Subsequently, we devised an image processing algorithm for real-time key point localization and DBH measurement, leveraging historical data to fine-tune current frame assessments. This system facilitates real-time image data upload via wireless LAN for immediate host computer processing. We validated our approach on seven sample plots, achieving bbAP50 and segAP50 scores of: 90.0%(+3.0%), 90.9%(+0.9%), respectively with the improved YOLOv5-seg model. The method exhibited a DBH estimation RMSE of 17.61∼54.96 mm (R2=0.937), and secured 78% valid DBH samples at a 59 FPS. Our system stands as a cost-effective, portable, and user-friendly alternative to conventional forest survey techniques, maintaining accuracy in real-time measurements compared to SFM- and LiDAR-based algorithms. The integration of WLAN and its inherent scalability facilitates deployment on Unmanned Ground Vehicles (UGVs) to improve the efficiency of forest inventory. We have shared the algorithms and datasets on Github for peer evaluations.

show abstract

Section: Data Evaluationmentioning

confidence: 99%

An Advanced Software Platform and Algorithmic Framework for Mobile DBH Data Acquisition

Zhang,

Liang,

Tong

et al. 2023

Forests

View full text Add to dashboard Cite

show abstract

“…Deep learning technology is a subfield of artificial intelligence, which has proven well performance in semantic segmentation using deep networks. [21][22][23] The task of crack detection is treated as a semantic segmentation, in which the user provides crack images and then obtains information about cracks. 24) Recently, many crack detection methods using deep network have adopted in different case studies due to the high accuracy and good robustness.…”

Section: Study Of Crack Detection Networkmentioning

confidence: 99%

Evaluation of Deep Network-based Methods for Crack Detection of Iron Ore Green Pellet

et al. 2022

View full text Add to dashboard Cite

show abstract

“…The characterization of a crack pattern and divided cell morphology has also been investigated, and not only surface patterns but also three-dimensional morphology have been explored [7] [8]. In addition to laboratory-based studies, large-scale and long-term experiments have been conducted [9], and the introduction of information technology such as deep learning has been in progress [10]. Concerning the modeling and mechanism of crack formation, the effects of material properties and geometry of the specimen correlated with tensile stress and strain have been investigated [11] [12], and these studies have been performed not only in geology but also in other fields such as chemical engineering [13].…”

Section: Introductionmentioning

confidence: 99%

Simulation of Crack Pattern Formation Due to Shrinkage in a Drying Material

Uehara¹

2023

OJMSi

View full text Add to dashboard Cite

Crack patterns observed in nature have attracted the interest of researchers in various fields, and the mechanism of the pattern formation has been investigated. However, the phenomenon is very complicated, and many factors affect the process. Therefore, we are motivated to construct a general simulation code with a simple algorithm. In this study, crack pattern formation due to shrinkage caused by the drying of a wet material was simulated. The process was simplified as follows: tensile force is generated in the model, and a crack is generated when the tension exceeds a critical value. The tensile forces in the x and y directions are independently evaluated. A crack propagates perpendicular to the tension until it reaches another crack or a boundary. Based on this modeling, simulations with a two-dimensional square domain were performed. Consequently, a cross-divided pattern was generated. Assuming zigzag crack propagation, more realistic patterns were obtained. The effects of the boundary and domain size were also considered, and various characteristic patterns were obtained. Furthermore, the orientation dependency was simulated, and 45˚ declined patterns and rectangularly divided patterns were generated. The model presented in this study is very simplified and is expected to be applicable to various objects.

show abstract

Deep learning based approach for the instance segmentation of clayey soil desiccation cracks

Cited by 22 publications

References 38 publications

An Advanced Software Platform and Algorithmic Framework for Mobile DBH Data Acquisition

An Advanced Software Platform and Algorithmic Framework for Mobile DBH Data Acquisition

Evaluation of Deep Network-based Methods for Crack Detection of Iron Ore Green Pellet

Simulation of Crack Pattern Formation Due to Shrinkage in a Drying Material

Contact Info

Product

Resources

About