We investigate methods for detecting and mapping features in landscapes using very high spatial resolution imagery (<1m pixel size). The focus of this line of research is to take advantage of imagery sources such as UAV cameras and the US National Agricultural Imagery Program (NAIP).

We applied optimized Geographic Object-Based Image Analysis (GEOBIA) approach for ingesting, processing, and classifying imagery. We reduce the spectral dimensionality of imagery using principal component analysis (PCA), texture analysis, and edge detection. Objects created through image segmentation are then used to implement a random forest algorithm for classification with minimal post-processing corrections.

We have applied these methods in projects addressing plant species mapping, species habitat distribution, deforestation mapping and urban tree cover mapping. 

For more information, contact the GST lab's PI Dr. Carlos Portillo-Quintero at carlos.portillo@ttu.edu.