Locating Cave Entrances Using Lidar-Derived Local Relief Modeling

Abstract

Lidar (Light detection and ranging) scanning has revolutionized our ability to locate geographic features on the earth’s surface, but there have been few studies that have addressed discovering caves using this technology. Almost all attempts to find caves using lidar imagery have focused on locating sinkholes that lead to underground cave systems. As archaeologists, our work in the Chiquibul Forest Reserve, a heavily forested area in western Belize, focuses on locating potential caves for investigation. Caves are an important part of Maya cultural heritage utilized by the ancient Maya people as ritual spaces. These sites contain large numbers of artifacts, architecture, and human remains, but are being looted at a rapid rate; therefore, our goal is to locate and investigate as many sites as possible during our field seasons. While some caves are entered via sinkholes, most are accessed via vertical cliff faces or are entered by dropping into small shafts. Using lidar-derived data, our goal was to locate and investigate not only sinkholes but other types of cave entrances using point cloud modeling. In this article, we describe our method for locating potential cave openings using local relief models that require only a working knowledge of relief visualization techniques. By using two pedestrian survey techniques, we confirmed a high rate of accuracy in locating cave entrances that varied in both size and morphology. Although 100% pedestrian survey coverage delivered the highest rate accuracy in cave detection, lidar image analyses proved to be expedient for meeting project goals when considering time and resource constraints.

Using Deep Learning and Hough Transformations to Infer Mineralised Veins From Lidar Data Over Historic Mining Areas

“…However, they conclude that applications using LiDAR data, either applied directly to the point cloud or using an image-like gridded representation, are less frequently studied than applications based on optical data. Many of the published LiDAR based remote sensing applications come from archaeology, where LiDAR is a widely used data source for both human interpretation of heritage landscapes (Bewley et al 2005, Hesse 2010and Moyes and Montgomery 2019 and semi-automated site detection based on template matching or traditional "shallow" machine learning methods (Freeland et al, 2016, Sevara et al, 2016and Guyot et al, 2018. Despite the often simple geometry of the sites to be detected, the accuracies of these methods generally cannot approach human levels (Verschoof-van der Vaart and Lambers, 2019).…”

Bringing Lunar LiDAR Back Down to Earth: Mapping Our Industrial Heritage through Deep Transfer Learning

“…Only pits within a section of the dataset were ground truthed as shown in Figure 10c. 2 The Yorkshire dataset was exported for testing without human generated labels.…”

“…Airborne LiDAR systems are an increasingly valuable tool for locating, visualizing and understanding cultural heritage sites. The ability to perceive subtle depressions and patterns in the landscape uncoupled from photometric representations has led to discoveries ranging from additional monuments at Stonehenge [1] to Mayan cave dwelling entrances in Belize [2]. These are examples of the more traditional uses of LiDAR in archaeology, where data over a small area is visually analyzed through an experience and knowledge based process to obtain a detailed understanding of a landscape, usually by displaying the LiDAR data as a hillshaded image [3].…”

“…Analyzing these volumes of data efficiently by human operators is extremely challenging. For example, the English Heritage National Mapping Program (primarily aerial image interpretation) achieves a coverage rate of approximately 1 km 2 per person per day; this project has been running for over 20 years employing on average 15-20 staff and had covered an area of 52,000 km 2 by 2012, in contrast, the Baden-Württemberg study, whilst still a primarily manual approach, took advantage of automated processing where possible, allowing an estimated coverage rate of over 35,000 km 2 by a single operator in six years [7]. These two projects are not directly comparable, as the quality and accuracy of their results varies greatly along with the data types analyzed [3], but it is an indication of the speed advantages gained from integrating automated processes into an analysis workflow.…”

Earth Observation, Remote Sensing, and Geoscientific Ground Investigations for Archaeological and Heritage Research

“…The contribution by Moyes & Montgomery [30] adds a further proof of the usefulness of this technology to explore Maya lowlands and other tropical regions, where dense vegetation usually prevents archaeologists from conducting extensive surveys or, at least, makes this type of archaeological survey less cost-effective. In particular, the authors describe a method for locating potential cave openings using local relief models that require only a working knowledge of relief visualization techniques.…”