Submitted to: American Society for Photogrammetry and Remote Sensing Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 12/1/2005
Publication Date: 5/1/2006
Citation: Laliberte, A.S., Rango, A., Fredrickson, E.L. 2006. Separating green and senescent vegetation in very high resolution photography using an intensity-hue-saturation transformation and object based classification. In: Proceedings of the American Society for Photogrammetry and Remote Sensing Annual Conference, May 1-5, 2006, Reno, Nevada. 2006 CDROM. Interpretive Summary: No interpretive summary required.
Technical Abstract: In arid regions of the southwestern US, grass cover is typically a mixture of green and senescent plant material. It is important that both types of vegetation can be quantified for land management purposes and for assessing the nutritional value of grasses. Traditional ground sampling procedures are commonly used but are time consuming. Our goal was to develop an image analysis approach for separating and quantifying green and senescent grasses in the same plot using very high resolution ground photography. The study was conducted in New Mexico at the Jornada Experimental Range (JER), operated by the USDA Agricultural Research Service, where ongoing research is aimed at determining the relationship between ground-based observations and remotely sensed data. We used an eight megapixel digital camera to acquire ground photography from a height of 2.8 m above ground for fifty plots in a stratified random sample approach. The area had high vegetation variability and each plot covered 2.5 m x 3.5 m. Preliminary studies have shown that a transformation from the RGB (red, green, blue) color space to the IHS (intensity, hue, saturation) color space was advantageous for separating green and senescent vegetation. We used an object-based image analysis approach to classify the images into soil, shadow, green vegetation, and senescent vegetation. A multitude of spectral, spatial and textural features were available for analysis, and the most suitable features were determined with a feature space optimization method. Green and senescent vegetation were best separated best by using the hue band, while the saturation band best differentiated between soil and senescent vegetation. The highest classification accuracies for the four classes were achieved by using IHS bands and omitting RGB bands. Ongoing research is designed to relate the results to ground collected information and to aerial photography and QuickBird imagery.