Kelly process drone images for irrigation prediction metrics

Who is the user Kelly is an irrigation analyst that flies drones to determine how well plants are growing with his water regimen at various times over the growing season. Averages approximately 7 flights per season.

What is the user doing After the drone photo capture, each image has its ground control points identified by hand, then OpenDroneMap is used to create a mosaic, finally each plot has its greenness calculated to determine how well the water regimen is taking effect.

The problem Kelly's work computer runs OpenDroneMap to create the mosaic image. This can take until the next day to finish and takes over his machine potentially throttling throughput.

The solution Provide a means to load the captured images and the ground control point file to be processed by OpenDroneMap at a computing center to provide faster turnaround times and free his machine.

1) Upload drone data and control points to a secure environment
2) Automatically perform the following steps
    a) Create a mosaic image using OpenDroneMaps
    b) Provide a location for storing the resulting image for later downloading

Additional Desired Steps 1) Automatically extract plots from the mosaic image 2) Calculate the greenness per plot using Kelly's algorithm 3) Provide automated ground control point detection

Additional Benefits

By transferring the work of mosaicing the image to UA there's the ability to try different mosaicing software, as it becomes available to the pipeline, with minimal impact on Kelly's machine.
There is already one greenness calculation available. Kelly would be able to choose between different algorithms to determine which is best suited for his needs.
Kelly would be able to submit jobs at a faster pace than he would be able to process on his machine.

terraref / drone-pipeline

Kelly process drone images for irrigation prediction metrics #33