Skip to main content
ARS Home » Research » Publications at this Location » Publication #167035
Title: COTTON CROP SPECTRAL IMAGING ANALYSIS: A WEB BASED HYPERSPECTRAL SYNTHETIC IMAGERY SIMULATION SYSTEM

Author
item ALARCON, VLADIMIR - MISS STATE UNIV
item Sassenrath, Gretchen

Submitted to: Proceedings of SPIE
Publication Type: Proceedings
Publication Acceptance Date: 7/17/2004
Publication Date: 8/2/2004
Citation: Alarcon, V.J., Sassenrath Cole, G.F. 2004. Cotton crop spectral imaging analysis: a web based hyperspectral synthetic imagery simulation system. In: Remote Sensing and Modeling of Ecosystems for Sustainability, Gao, W. and Shaw, D.R., Eds. Proceedings of SPIE. Conf. Vol. 5544, 178-185.

Interpretive Summary: Remotely sensed data is sometimes used to generate spectral 'libraries' of reflectance information about specific parameters of interest. Previously, these libraries have been of limited accessibility due to the large size of the files, and the difficulty in making the data readily available. The research presented in this manuscript is designed to explore changes in remote imagery from cotton crop canopies with drought stress. We are examining the individual components that contribute to the overall canopy spectral reflectance. Previous research detailed a method of analyzing and visualizing the reflectance spectra by randomly mixing the reflectance from individual components into a synthetic image of the crop canopy. This method allows us to explore the individual factors of the spectra, their contribution to the overall spectral signature, and the impact of environmental conditions and drought stress exposure on each factor. The work presented here expands on that previous work, and presents a Java interface to the synthetic imagery simulation. The Java interface is easy to use and allows data to be readily available through a web-based interface. In addition to improving the analysis and visualization of the reflectance spectra, the method developed in this research will allow web-based hyperspectral libraries for greater accessibility of images for public use.

Technical Abstract: The development of spectral libraries for specific vegetation species and soils is useful for identifying different physiological or physical-chemical characteristics. Usually, spectral libraries are provided as a data-base add-in of current commercial software used for analyzing hyperspectral imagery. The use of those databases requires installation of the software in the user's machine for either visualizing or using the spectral libraries. There are also spectral libraries available on the web but the data is static and partitioned by spectrum of vegetation or soil because the size of the files of actual hyperspectral images precludes it's publication on the web. In this paper, a web-based simulation environment for generating hyperspectral synthetic imagery of cotton plots is presented. The system was developed using Java and is based on a previous synthetic imagery program (Sassenrath et al., 2003). The mathematical and numerical formulation of the model is briefly sketched. The core computing components of the simulation environment were written in C for their computational efficiency. The emerging Java Native Interface (JNI) technique and standard Java techniques were used to design a user-friendly simulator. The simulation system provides interactive user control and real time visualization of the resulting hyperspectral image through standard web browsers. It shows potential for providing web-based hyperspectral libraries, in the form of images, for public use.