Semantic Segmentation of Airborne LiDAR Data in Maya Archaeology

Bundzel, Marek; Jaščur, Miroslav; Kováč, Milan; Lieskovský, Tibor; Sinčák, Peter; Tkáčik, Tomáš

doi:10.3390/rs12223685

Cited by 29 publications

(31 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Until now, the focus has been limited to a handful: mound structures (burial mounds [59,[62][63][64][65][66][67][68], charcoal kilns [69,70], and shell-rings [71]), pit structures (hunting system [72]; ore extraction pits [72,73], and bomb craters [74]), and linear sunken structures (paths [75,76], ditches [77], and mining shafts [73]). There is a recent trend of targeting complex features [78] and multiple feature types (multi-class archaeological object detection [77,[79][80][81][82][83][84]), but complex archaeological landscapes imbedded in a complex terrain with ample anthropogenic influence remains challenging [80,81].…”

Section: Archaeological Interpretation (31-35)mentioning

confidence: 99%

Documentation of Archaeology-Specific Workflow for Airborne LiDAR Data Processing

Lozić

Štular

2021

Geosciences

View full text Add to dashboard Cite

Airborne LiDAR is a widely accepted tool for archaeological prospection. Over the last decade an archaeology-specific data processing workflow has been evolving, ranging from raw data acquisition and processing, point cloud processing and product derivation to archaeological interpretation, dissemination and archiving. Currently, though, there is no agreement on the specific steps or terminology. This workflow is an interpretative knowledge production process that must be documented as such to ensure the intellectual transparency and accountability required for evidence-based archaeological interpretation. However, this is rarely the case, and there are no accepted schemas, let alone standards, to do so. As a result, there is a risk that the data processing steps of the workflow will be accepted as a black box process and its results as “hard data”. The first step in documenting a scientific process is to define it. Therefore, this paper provides a critical review of existing archaeology-specific workflows for airborne LiDAR-derived topographic data processing, resulting in an 18-step workflow with consistent terminology. Its novelty and significance lies in the fact that the existing comprehensive studies are outdated and the newer ones focus on selected aspects of the workflow. Based on the updated workflow, a good practice example for its documentation is presented.

show abstract

Section: Archaeological Interpretation (31-35)mentioning

confidence: 99%

Documentation of Archaeology-Specific Workflow for Airborne LiDAR Data Processing

Lozić

Štular

2021

Geosciences

View full text Add to dashboard Cite

show abstract

“…Overall Accuracy (OA) [99] Percent Correct Classification [100] Pixel Accuracy [101] TP+TN TP+TN+FP+FN…”

Section: Relation To Traditional Rs Measuresmentioning

confidence: 99%

Accuracy Assessment in Convolutional Neural Network-Based Deep Learning Remote Sensing Studies—Part 1: Literature Review

2021

View full text Add to dashboard Cite

Convolutional neural network (CNN)-based deep learning (DL) is a powerful, recently developed image classification approach. With origins in the computer vision and image processing communities, the accuracy assessment methods developed for CNN-based DL use a wide range of metrics that may be unfamiliar to the remote sensing (RS) community. To explore the differences between traditional RS and DL RS methods, we surveyed a random selection of 100 papers from the RS DL literature. The results show that RS DL studies have largely abandoned traditional RS accuracy assessment terminology, though some of the accuracy measures typically used in DL papers, most notably precision and recall, have direct equivalents in traditional RS terminology. Some of the DL accuracy terms have multiple names, or are equivalent to another measure. In our sample, DL studies only rarely reported a complete confusion matrix, and when they did so, it was even more rare that the confusion matrix estimated population properties. On the other hand, some DL studies are increasingly paying attention to the role of class prevalence in designing accuracy assessment approaches. DL studies that evaluate the decision boundary threshold over a range of values tend to use the precision-recall (P-R) curve, the associated area under the curve (AUC) measures of average precision (AP) and mean average precision (mAP), rather than the traditional receiver operating characteristic (ROC) curve and its AUC. DL studies are also notable for testing the generalization of their models on entirely new datasets, including data from new areas, new acquisition times, or even new sensors.

show abstract

“…Another potential avenue for research is the use of semantic segmentation models that would be able to segment objects on a pixel level (see for instance Bundzel et al 2020;Kazimi, Thiemann & Sester 2019). While bounding boxes are adequate for the localisation of objects, segmentation might offer additional information concerning their size, coverage, etc.…”

Section: Further Improvementsmentioning

confidence: 99%

Implementing State-of-the-Art Deep Learning Approaches for Archaeological Object Detection in Remotely-Sensed Data: The Results of Cross-Domain Collaboration

Olivier

Vaart

2021

Journal of Computer Applications in Archaeology

View full text Add to dashboard Cite

The ever-increasing amount of remotely-sensed data pertaining to archaeology renders human-based analysis unfeasible, especially considering the expert knowledge required to correctly identify structures and objects in these type of data. Therefore, robust and reliable computer-based object detectors are needed, which can deal with the unique challenges of not only remotely-sensed data, but also of the archaeological detection task.In this research -across-domain collaboration between archaeology and computer science -the latest developments in object detection and Deep Learning -for both natural and satellite imagery -are used to develop an object detection approach, based on the YOLOv4 framework, and modified to the specific task of detecting archaeology in remotely-sensed LiDAR data from the Veluwe (the Netherlands). Experiments show that a general version of the YOLOv4 architecture outperforms current object detection workflows used in archaeology, while the modified version of YOLOv4, geared towards the archaeological task, reaches even higher performance. The research shows the potential and benefit of cross-domain collaboration, where expert knowledge from different research fields is used to create a more reliable detector.

show abstract

Semantic Segmentation of Airborne LiDAR Data in Maya Archaeology

Cited by 29 publications

References 34 publications

Documentation of Archaeology-Specific Workflow for Airborne LiDAR Data Processing

Documentation of Archaeology-Specific Workflow for Airborne LiDAR Data Processing

Accuracy Assessment in Convolutional Neural Network-Based Deep Learning Remote Sensing Studies—Part 1: Literature Review

Implementing State-of-the-Art Deep Learning Approaches for Archaeological Object Detection in Remotely-Sensed Data: The Results of Cross-Domain Collaboration

Contact Info

Product

Resources

About