Wide Area Survey and Targeting

From UAS Operations for Emergency Management
Jump to: navigation, search

Effectively employing a SUAS as part of a wide area search requires a methodical and disciplined approach to cover the most area in the shortest amount of time. In addition, to minimize the time required, this workflow is resource intensive - more people working as analysts is better.

Two workflows are illustrated here. The first, Georectified Targeting is the preferred method, and requires a computer running Agisoft Photoscan Professional as well as personnel trained to utilize it. The method is preferred because it produces better quality (more likely to be valid) targets and more precise coordinates for those targets.

The second workflow, EXIF-Only Targeting, can produce valid targets, but the coordinate precision will be lower, and they will not be reviewed from multiple camera angles. Whenever practical, empty the Georectified Target method.

As targets are developed (using either method), the SAUS team must distribute this information to the Operations Branch, Air Operations Branch, or other responsible party in the response structure. This is the entire point of the SUAS operations! Targets are distributed using the Target List in either .CSV or .KML formats.

Georectified Targeting (Preferred)

The most effective targeting method requires a functional version of Agisoft Photoscan Pro as well as Google Earth Pro. it provides accurate coordinates for objects detected (targets) by the SUAS photo survey.

Between Steps 4 and 8, Agisoft Photoscan Pro will not be available for squint or image analysis. If critical targets are identified in raw images, a short version of the EXIF-Only Targeting (Steps 5 and 6) can be utilized.

  1. Capture nadir still images (perform a SUAS photo survey)
    1. Utilize a mapping software package such as Mission Planner, Maps Made Easy Map Pilot, Drone Deploy, etc.
    2. If the images will be used to create an orthomosaic, use standard overlap settings
    3. If the images will just be used for targeting (the squint), then reduce overlap. Note that some overlap is required. For reference, 30%-40% forward overlap and side lap is sufficient for SAR.
  2. Load captured photos onto processing laptop.
  3. Import captured photos into Agisoft Photoscan Pro on processing laptop.
    1. Align Photos in Photoscan Pro
    2. Build Mesh in Photoscan Pro. NOTE: This is not a complete processing chain in Photoscan Pro. This is the minimum required for targeting processing.
  4. While photos are being processed in Photoscan Pro….
  5. Load Captured Photos onto squint devices - additional laptops (not for processing), tablets, etc.
  6. Divide workload among available personnel with squint devices. The manual squint will be a parallel process. Personnel will work through assigned photos on different devices. Additionally, the ortho processing will continue as a separate process.
  7. Perform the targeting squint on as many devices as possible.
    1. Identify potential targets in unprocessed photos
    2. Note photo numbers for later use.
  8. When Agisoft Photoscan processing is complete (from Step 3)…
  9. Export Photoscan Pro project onto external HD.
  10. As targets are identified in unprocessed photos…
    1. The target is located in the same photo in the Photoscan interface.
    2. The target is flagged using a “Marker” in Photoscan. Right-click on the target visible in a photo. Select ‘Create Marker’. In the Markers window in the right hand References pane, select the new Marker and rename using a logical naming scheme (such as invest 1, invest, 2, invest 3, etc.).
    3. In the Reference pane for Markers, select, “View Estimated”.
    4. The estimated coordinates of the target is displayed next to the target Marker name.
  11. Evaluate the target for value and applicability…
    1. Right click on the marker name in the Reference pane.
    2. Select “Filter Photos by Markers”. Only photos containing the possible target Marker will be displayed in in the Photos pane. Each photo containing the Marker will have the Marker location indicated.
    3. Examine each photo containing the Marker to determine the class and identity of the object indicated. This process will provide a variety of different angles to examine the possible target.
  12. If the analyst determines the target to be valid or of concern, add it to the Target List.

EXIF-Only Targeting (Resource Constrained)

In situation where Agisoft Photoscan is not available or not practical, targeting can be accomplished suing raw (unprocessed) SUAS survey photos. This method is much less precise than Georectified Targeting and does not allow for multi-angle/multi-image review of potential targets. This can result in a higher rate of false positive targets.

An abbreviated version of this Targeting method (Steps 5 and 6) can be utilized while Georectified Targeting is processing images or mesh in PhotoScan Pro. This Quick Targeting should be verified as soon as possible.

  1. Capture nadir still images (perform a SUAS photo survey)
    1. Utilize a mapping software package such as Mission Planner, Maps Made Easy Map Pilot, Drone Deploy, etc.
    2. If the images will be used to create an orthomosaic, use standard overlap settings
    3. If the images will just be used for targeting (the squint), then reduce overlap. Note that some overlap is required. For reference, 30%-40% forward overlap and side lap is sufficient for SAR.
  2. Load captured photos onto analysis/squint laptops or devices.
  3. Divide workload among available personnel with squint devices. The manual squint will be a parallel process. Personnel will work through assigned photos on different devices.
  4. Perform the targeting squint on as many devices as possible.
    1. Identify potential targets in raw photos
    2. Note photo numbers of images that contain potential targets.
  5. As targets are identified in photos…
    1. Utilizing an EXIF data extractor (including OS photo viewers) to determine the latitude and longitude of the center point of the photo.
    2. Locate that coordinate in Google Earth Pro.
    3. Using a visual comparison, determine the approximate of the target using Google Earth imagery.
  6. Evaluate the target for value and applicability… If the analyst determines the target to be valid or of concern, add it to the Target List.

Creation and Distribution of the Target List

The Target List is the primary product of the SUAS SAR Workflow. It provides information about potential targets obtained from SUAS surveys of the search area in an easy to consume format.

The standard Target List is a Excel document saved in .CSV format. This file format is usable in Excel or any other spreadsheet program or text editor. Further the Target List .CSV format can be imported to Google Earth Pro or other GIS software packages.

The standard column are as follows: “Target_ID”, “Latitude”, “Longitude”, “Elevation_Estimate”, “Category”, “Description”, “Date”.

Target_ID
Unique identifier developed by the analyst adding targets to the Target List.
Latitude
Latitude of the target mark, utilizing Decimal Degrees (DD) format*
Longitude
Longitude of the target mark, utilizing Decimal Degrees (DD) format*
Elevation_Estimate
Text representation of the elevation of the target above the ground. If the target appears to be located on the ground, stated as “Surface”.
Category
If searching for multiple objects or classes of objects, Category can be used to differentiate targets. Category is flexible and assigned by the analyst.
Description
Text description (brief!) of the target. May include offset info if the target i snot located directly at the estimated coordinates.
Date
Date the target was developed.

The target list can be distributed as the basic .CSV file, or as a Google KML format file. To create a .KML:

  1. Open Google Earth Pro
  2. Select File -> Import
  3. Select the Target List .CSV file and select “Open”
  4. When prompted to apply a style template, select “Yes”.
  5. If there is an existing style template to use, select “Use existing template” and then select the template. Click “OK”. Proceed to Step 12.
  6. Select “Create new template” and click “OK”
  7. Name is the data that appears next to the Target on the map. Select “Target_ID” in the “Set name field”.
  8. Select “Icon” in the top row.
  9. Select “Use same icon for all features” and then select the icon in the pulldown box.
  10. Click “OK”
  11. If promoted to save the template, do so. You can then reuse that in Step 5.
  12. The .CSV is imported and targets are plotted in the main Google Earth map display.
  13. In Google Earth Pro, locate the newly imported .CSV file in the Places pane. By default when imported, a file is placed in the Temporary Locations folder at the bottom.
  14. Right click on the imported CSV file layer.
  15. Select “Save Place As…”
  16. Navigate to the preferred save location. Enter an appropriate file name.
  17. Select “Kml (*.kml)” in the Files of type dropdown box.
  18. Click “Save”

The KML version of the Target List can be distributed via USB drive, email, etc. It can be consumed by Google Earth Pro. Basecamp, and a variety of other GIS and/or mapping programs.

Retrieved from "http://uas.cdrp.net/index.php?title=Wide_Area_Survey_and_Targeting&oldid=189"