Narrow-linear and Small-area Forest Disturbance Detection and Mapping from High Spatial Resolution Imagery: a Framework

Over the past decade, widespread disturbance has brought a large amount of narrow-linear and small-area disturbance features (e.g., ATV trails, seismic lines, pipelines/powerlines, well sites, forest roads and cut blocks) to forest areas in the Foothills Region of Alberta, Canada. This issue has prompted research into finding appropriate data and methods for mapping these narrow-linear and small-area disturbance features in order to examine their impacts on wildlife habitat. Recent released high-resolution remote sensing data and introduced image processing methods have the potential to update detail information on rapidly increased small forest disturbance. In this paper, we first described the characteristics of small forest disturbances and presented the nature of problems. Then we presented a framework for detecting and extracting narrow-linear and small-area forest disturbance features. Following the framework, we provided an overview of the types of imagery being used for small-feature extraction, discussed the methods that have potential in support of mapping small disturbance from high resolution imagery, and included experimental results in terms of mapping accuracy and completeness. The challenges and future directions for detection and mapping of small-scale disturbance are identified and synthesized at the end.