I increasingly see references to “mapping” this or that simply based on capturing image data of the area in question. Mapping is much more than image data acquisition, it is the interpretation of such data and graphical representation of meaningful information by the map’s creator. It is the process of turning data into information.
The amount of available image data is growing tremendously – from the number of satellite sensors available to the increasing popularity of data acquisition with drones; we are generating more and more data. But, what are we doing with all these beautiful data products?
I recently was able to spend a week in southern Louisiana with Dr. Whitney Broussard from JESCO Environmental. Dr. Broussard has helped JESCO to develop a new business area based on the combination of hyperspatial UAS data products from a Trimble UX5 and object-based image analysis (OBIA) routines created using Trimble eCognition to map sensitive marshland regions. The techniques developed and implemented by JESCO are featured in the September 2018 edition of Point of Beginning (POB) – “Marshland Mapping uses Drone and Data” and even made the cover story.
Dr. Whitney Broussard is often confronted with explaining the importance of mapping the marshland areas. He spelled it out to us pretty clearly, noting that Louisiana has lost nearly 1,900 square miles (4,921 km2) of land since the 1930s and if preventative actions are not taken, the coastal area could lose another 4,100 square miles (10,619 km2) of land in the next 50 years. As a reaction, the state has “established 390 coastal reference monitoring stations (CRMS) to measure the health of the coastal marsh vegetation and the impacts of erosion”. But these areas are difficult to monitor.
One of the main challenges when mapping coastal marshlands has always been accessibility, JESCO’s project area “was 4 kilometers south of the nearest terra firma”. Broussard goes on to say that “accessing it via airboat and on foot can disrupt the vegetation you are trying to protect”. In addition, traditional remote sensing methods, based on satellite and aerial imagery, simply do not provide the level of detail required by the project. “Traditional aerial surveys and image processing aren’t fine enough to precisely classify the land/water boundary and varied vegetation” say Broussard. They have a tendency to overestimate land cover percentages
The solution developed by Dr. Broussard and JESCO – pairing UAS data products with Trimble’s eCognition Developer software – has led to “new business revenue streams” and is helping to “redefine the business of vegetation mapping for state and local authorities altogether”.
“The magic of eCognition is in its segmentation,” says Broussard. “Once you have set your parameters within the segmentation process, eCognition uses those parameters to group pixels into units that share similar attributes and categorize them. It mimics the human brain’s process of identifying objects through pattern recognition. Using a rule set ensures that you are capturing the objects you want classified because the software won’t deviate from the rules. That’s why OBIA is able to methodically and repeatedly do something that humans can’t do.”
When compared to the state’s Coastal Protection and Restoration Authority (CPRA) 2012 data set, JESCO’s results identified 100 percent of the vegetation types and calculated plant heights to within 88-94 percent. The resulting land cover maps were “strikingly more detailed” since working “with drones and OBIA technology, instead of producing point data every few hundred feet, we produce models every few centimeters,” says Broussard.
The reaction from the state to these new high accuracy marshland maps has been so positive that JESCO has branched out into an entirely new business area. The survey company C.H. Fenstermaker hired JESCO to survey and classify a wetlands mitigation bank at the Rockefeller Wildlife Refuge (RWR), a 71,500-acre wildlife and fisheries refuge.
Due to the nature of traditional marshland survey techniques, RWR was concerned that repeated measurements of the vegetation with this methodology could be harmful to the wetlands health – They wanted a more non-invasive and accurate approach. JESCO’s team was sent into fly and map the wetland in November 2017. The UAS flights generated 4,899 images with 2.5 cm GSD that were used to create a DSM and orthomosaic as input for Broussard’s eCognition solution. eCognition delineated the land/water boundaries and classified four target classes: grass (Spartina sp), reed (Phragmites australis), shrub/scrub and impervious in less than 2 hours.
“A detailed, OBIA-based vegetation map gives managers a meaningful measurement of their wetlands environment,” says Broussard. “It also provides a highly accurate map of their wetland acreage–rather than just an estimation–which they can use in their required reporting to authorities.”