ASSESSING NATURAL RESOURCES WITH THE VISNIRMIR DIGITAL VIDEO IMAGING SYSTEM

摘要

The visible/near-infrared/mid-infrared (VISNIRMIR) digital video imaging system acquires data in the blue (447-455 nm), green (555-565 nm), red (625-635 nm), red edge (704-716 nm), near-infrared (814-826 nm), and mid-infrared (1631-1676 nm) regions of the electromagnetic spectrum. This paper discusses application of the system as a tool for assessing natural resources. For this study, a rangeland area and a wetland area were used as test sites. Conventional color and false color images were created and were qualitatively assessed to determine their usefulness for separating various land-cover types within the areas of interest. The imagery was also subjected to an unsupervised classification to evaluate the potential of the system as tool for land-cover mapping. Composite imagery provided valuable information for qualitative assessment of the study sites. For the rangeland site, the best map accuracies were obtained using a broad land-cover mapping scheme. American lotus (Nelumbo lutea) was classified accurately for the wetland site (user's and producer's accuracies greater than 90%); the other land-cover types were mapped with low (<60%) to intermediate accuracies (60 to 84%). These preliminary results indicate that the mapping capabilities of this system are site and cover type specific.