Weed ID with Remote Sensing

Using hyperspectral remote sensing to identify weeds and vegetation types is being addressed in several CalSpace projects. Ongoing efforts to show others how to apply remote sensing include workshops and tutorials being developed from and for these training sessions.

As an introduction to identifying weeds with remote sensing, here is an outline of work in progress. These Flash movies, based on talks given for a SERDP workshop held at CalSpace, are an outline of concepts covered in the tutorials. Please download a PDF version of the tutorial. In it you will find some background in field technique and a step by step approach to analyizing the sample data.

To view these movies, you will need to install Macromedia Flash Player (if you do not already have it). It is available for free at the Macromedia website. Because the program loads one page at a time, successive downloads are kept to a few hundred kb and your speed-of-access should be acceptable, even if you use a dial-up modem instead of higher speed cable, T1, or DSL.

Not all of the topics have audio and presentation notes at this time, but they will soon. If you click on Maximum Likelihood, you will hear Ed Tom lecture. Note that you can start and stop talking with the control buttons below his picture and talk outline, send him an email, and view slide notes - all at the same time the video plays.

Using Hyperspectral Remote Sensing to ID Weeds

Introduction
Spectra
Imagery - Preprocessing
Imagery - Classifying
Imagery - Reducing dimensionality
Imagery - Analyzing at the subpixel level
Imagery - Verifying classification