Determining a Confidence Interval for Repeat sUAS Imagery to Assess Vegetation Cover Change
Topics: UAS / UAV
, Remote Sensing
,
Keywords: UAV, Drone, UAS, Vegetation Mapping, Change Detection, Coastal, Remote Sensing
Session Type: Virtual Paper
Day: Thursday
Session Start / End Time: 4/8/2021 03:05 PM (Pacific Time (US & Canada)) - 4/8/2021 04:20 PM (Pacific Time (US & Canada))
Room: Virtual 13
Authors:
Grayson Morgan,
,
,
,
,
,
,
,
,
,
Abstract
Abstract
Coastal zones are dynamic regions of constant change with dune systems that are fortified and stabilized by vegetation. The importance of vegetation to the dune systems and coastal zones warrant regular monitoring to help determine the health of the plant system. Such monitoring is important for not only understanding the science of vegetation change but for coastal management challenges – particularly after catastrophic events, such as hurricane visits. A wealth of prior research has been conducted comparing various change detection techniques, as well as an analysis of those techniques in the coastal zone. However, small unmanned aerial systems (sUAS) remote sensing provides new potential for change detection as well as new challenges. With such high-resolution imagery being collected at lower altitudes, small ‘apparent’ changes in multidate imagery can be considered a ‘change’ by algorithms, even though there may be no true change on the ground. In this research we examine the creation and use of a confidence interval to help interpret vegetation change by using a post-classification change detection method. The confidence interval, determined using 10 flights over the same study area in a period of 2 hours, will be applied to evaluate ‘apparent’ changes in sUAS imagery captured before and after Hurricane Irma affected Harbor Island, SC in September 2017. The results of this research will indicate the effectiveness of using a priori determined confidence intervals to assess changes in other sUAS imagery over similar landscapes.