Case Study: Environmental Assessment Using LiDAR Data and Near-Ground Feature Detection Algorithms

‍

Image source

Client Confidentiality Statement: The client's identity for this case study is kept confidential to respect their privacy and proprietary interests. The information presented here is intended to highlight the data analysis process and outcomes of the project without disclosing the client's identity.

Project Overview:

The project spans across >400km² of the Goldfields area. The project aimed to perform data analysis and identify potential Malleefowl mounds within the defined Area of Interest (AOI). The project was delivered using the supply of Outline Globals (10cm resolution) imagery and raw LiDAR data.

‍

Data Analysis Process:

Outline Global supplied the LiDAR and imagery data to be analysed by Anditi. The process required the classification of terrain features into distinct categories including ground, vegetation, and other non-ground elements. The data was further processed to create a Digital Elevation Model (DEM) that incorporated potential mound-like features. The specialised near-ground feature detection algorithms developed by Anditi were then employed to identify potential Malleefowl mound sites within the AOI. These sites were ranked based on their level of certainty, considering factors such as mound intactness, vegetation obscuring, data gaps, and variability.

‍

Manual checks were conducted to validate the automated identification process to ensure accuracy. An orthophoto was used for comparison with the algorithm's ratings 1-3 to eliminate false positives. Anditi extracted various attributes for each identified mound, including location, height above sea level, mound radius, and mound height.

‍

Data Results:

The data analysis produced a comprehensive set of results. A classified point cloud, ground model, and orthophoto were provided. A total of 117 highly likely and definitive Malleefowl mounds were identified, alongside 144 findings that exhibited similar characteristics but could potentially be inactive mounds or vegetation. These results were recorded in a shapefile format with accompanying attributes.

1 rated mound example

‍

2 rated mound example

‍

Interpretation and Conclusions:

The analysis indicated the presence of several strong candidate mounds with ratings of 1 or 2, suggesting the potential presence of Malleefowl within the project area. The most promising mound candidates were predominantly situated within areas of taller vegetation where cover and shading were more prominent. Notably, a concentrated band of high-ranking (1, 2) mounds was observed on elevated terrain at the heart of the AOI. This suggested a possibly dense Malleefowl population, implying the environmental significance of this area.

‍

Challenges and Recommendations:

While the availability of an aerial image aided the verification process, certain limitations such as shadows and dense vegetation cover made the identification of false positives challenging. The study recommended field checks for mounds rated 1, 2, and 3 to confirm their nature. However, a significant portion of the mounds with a rating of 3 was likely unrelated to Malleefowl activities.

‍

Data Boundary Clarification:

It's important to note that data processing and analysis extended beyond the AOI. Therefore, an attribute field was introduced to distinguish whether a mound was identified within or outside of the AOI. This labelling approach helps segregate data originating from different areas.

‍

Summary:

This case study demonstrates the successful use of LiDAR data and Anditi's data processing engine to identify potential Malleefowl mounds within the defined project area. The analysis resulted in the identification of numerous candidate mounds, highlighting potential Malleefowl presence and environmentally significant areas. The study also emphasised the importance of combining automated algorithms with manual validation to ensure accuracy and reliability in the results.

‍

< Back to Case Studies