Capture Matching Roadmap

Roadmap

1. Manual Process

   1. Convergence algorithm Android complete since October 2021

      1. Also need testing under forest canopy

   2. Also collect step count and orientation on Android (but don’t use it yet)

      1. This data needs to be analyzed

   3. Captures are tagged per capture session

   4. We capture the full grower track path

      1. This data needs to be analyzed and leverages

   5. We also assume that we will know species

   6. Tools

      1. Capture Matching Tool, 80% complete

         1. Filter set: Session, species, distance

         2. Operator input: Landmarks, size

      2. QGIS - to identify capture records as ‘root captures’

   7. Should we include additional requirements on the planter

      1. Tagging trees, random dabs of paint

      2. Driven by resources of a given organization

   8. Profile all Android devices for GPS performance

      1. Generate list of suggested devices

2. Semi-automated process

   1. Using algorithms to improve and automate selection of potential match candidates

   2. Exploiting relative location between trees and autocorrelation of GPS errors

      1. Operator manually match ‘key trees’, or some percentage

      2. Solver would attempt to suggest a set (or candidates) of matches that uses this idea

   3. Spectral signature of forest lighting for a given capture may have additional information

   4. Automated capture matching becomes easier as total captures of a given tree increases (Manual also becomes easier)

3. Future Technologies

   1. L5 GPS

      1. Some testing done with L5, need more

      2. Should test with Garmin

      3. Request phones from a phone maker (garmin?) with this technology