Search

Science communication is increasing through the use of Unmanned Aerial Systems (UAS) or drones. Within the Arthur Temple College of Forestry and Agriculture at Stephen F. Austin State University (SFASU), UASs such as the DJI Phantom 4 Pro and Mavic Mini2 drones were used by students and faculty to study mistletoe, crepemyrtle and fire ants and then drone images were uploaded to iNaturalist, the largest repository for flora and fauna specimens to share with the scientific community and general public. The benefits of using a UAS is that nadir (directly above) images of the specimens increase the locational accuracy of each specimen compared to distance images acquired from a smartphone. By incorporating drones into course works at SFASU, faculty are increasing the technological abilities of students to communicate natural resource information to a greater audience as a citizen scientist. With ever increasing capabilities and lower cost, UAS are becoming a viable alternative to smartphones for communication of science, especially for iNaturalist. The ability to communicate science information and display images adds a dimension for the citizen scientist to use a UAS in teaching and information exchange while creating a well-rounded, better informed, and more employable student upon graduation.

The use of Unmanned Aerial Systems (UAS), also known as drones is increasing in geospatial science curricula within the United States. Within the Arthur Temple College of Forestry and Agriculture (ATCOFA) at Stephen F. Austin State University, Texas, seniors in the geospatial science program complete capstone projects to evaluate current geospatial technology to investigate complex ecological, social and environmental issues. Under the umbrella of a student initiated and designed senior project, students designed a study to estimate height of buildings with UAS data incorporating UAS data, LP360 and ArcScene programs, and Pictometry web-based interface. Results from a statistical analysis of the data confirm that geospatial science height estimation techniques can provide accurate estimates of height remotely. The independence of the students completing the project with UAS data for LP360 and ArcScene estimations, and utilizing Pictometry as an on-onscreen measuring tool, point to the need to integrate remote sensing, statistical analysis and synthesis of data into undergraduate geospatial science curricula. This reinforces the hands-on learning approach within ATCOFA and provides guidance to integrate the use of UAS in natural resource education.

CITYgreen Geographic Information Systems software was used to develop a campus wide cover type map for Stephen F. Austin State University in an environmental science landscape ecology course. The finding indicated an equal division of forest cover type compared to impervious surface of buildings and paved surface. Once the classification was completed, students chose an area for reforestation identified in CITYgreen, while raising funds for the purchase of trees for the project. Before completing the project, students reviewed tenets of landscape ecology, civic ecology education, and benefits of urban forestry. At the completion of the project, students reviewed service-learning aspects of campus beautification reflecting on making a difference, working outdoors, and using high end technology to complete a real-world environmental project incorporating partnerships and teamwork. The outcome demonstrates the benefits of applying ecological planning to complete an environmental project based on a perceived need within a campus setting.

The use of Unmanned Aircraft Systems (UAS) in undergraduate forestry education continues to expand and develop. Accuracy of data collection is an important aspect of preparation for "society-ready" foresters to meet the complex sustainable environment managing for ecological, social and economic interests. Hands-on use of a DJI Phantom 4 Pro UAS by undergraduates to measure the length and area of 30 linear features and areal features on Earth's surface were estimated. These measurements were compared (measured within the ArcMap 10.5.2 interface) to hyperspectral Pictometry imagery measured on the web-based interface and the Google Earth Pro interface. Each remotely estimated measurement was verified with the actual ground measurements and the methods compared. An analysis of variance, conducted on the absolute length errors resulting in a p-value of 0.000057, concluded that the three length estimating techniques were statistically different at a 95% confidence interval. A Tukey pair-wise test found that the remotely sensed DJI Phantom 4 Pro data was statistically less accurate than the Pictometry and Google Earth Pro data, while both of which were found to be not different statistically in terms of accuracy. The areal feature area measurements were not normally distributed and therefore tested for equal medians using a Kruskal-Wallis test. The test found that there was no significant difference between sample medians, indicating that all three methods of estimating area are statistically equal in accuracy. The results indicate that Pictometry and Google Earth Pro could both be used to accurately estimate linear feature lengths remotely in lieu of "in situ" linear measurements while all three remote sensing techniques can be used to accurately estimate areal feature areas remotely in lieu of "in situ" areal measurements.

Waterfowl populations monitoring is essential for wetland conservation. Lately, deep learning techniques have shown promising advancements in detecting waterfowl in aerial images. In this paper, we present performance evaluation of several popular supervised and semi-supervised deep learning models for waterfowl detection in aerial images using four new image datasets containing 197,642 annotations. The best-performing model, Faster R-CNN, achieved 95.38% accuracy in terms of mAP. Semi-supervised learning models outperformed supervised models when the same amount of labeled data was used for training. Additionally, we present performance evaluation of several deep learning models on waterfowl classifications on aerial images using a new real-bird classification dataset consisting of 6,986 examples and a new decoy classification dataset consisting of about 10,000 examples per category of 20 categories. The best model achieved accuracy of 91.58% on the decoy dataset and 82.88% on the real-bird dataset. [ABSTRACT FROM AUTHOR]