DroneMapper

DroneMapper is very pleased to announce a collaboration with DN2K and its partners in the development of value-added products for the precision agriculture market utilizing a combination of ground-based and aerial data collections.

DN2K operates a unique service called MyAgCentral (shown in the illustration), a subscription-based service for growers, retailers and service providers that offers precision agricultural data and integration to key software products for managing farms. Examples of data sets include heatmaps that graphically show key farming parameters such as: crop yield, fertilizer application, ground moisture, variable rate irrigation (VRI) and Normalized Difference Vegetation Index (NDVI). DroneMapper provides support to DN2K for terrestrial data visualization and processes geo-referenced orthos in the near infrared along with corresponding NDVI maps of the crop.

Please do not hesitate to contact DN2K (dn2k.com) and/or DroneMapper (dronemapper.com) for more detailed information and how we can support your precision farming needs.

DN2K/MyAgCentral UAS Press Release – March 5th, 2014

We’ve generated a heatmap showing some of the UAV hot spots around the globe. Data is a random sampling of 30,000 records from our imagery database.

We’ve implemented some basic data collection quality assurance and quality control tools inside the DroneMapper web interface. For each flight uploaded, we now generate an image footprint polygon layer. The polygons are oriented by obtaining the bearing between Latitude and Longitude points along the flight line, the size of the polygon is computed using the ground sample distance (GSD), sensor size and image size. The intention is to give you an idea of the overlap and imagery coverage over the area of interest, clearly showing large holes or gaps in the flight data collection. Other additions include basic EXIF geo-tag sanity checks such as unique Latitude and Longitude, consistent focal length, shutter speed / exposure time and missing altitude tags.

QA/QC Checks:

Image Footprint Polygons:

The image below represents an ideal data collection courtesy of Morgan from New Zealand. The flight is over mountainous terrain and has large areas of homogeneous ground cover. The image footprint polygons are visible and highlight the overall quality of the data collection. Morgan provides some additional information, and tips for other operators:

“If people ask you for tips and you don’t already know the following then feel free to tell them that flight was at 75% overlap, flown into and with the wind (so I got less crabbing of the air frame). I also triggered the camera based on distance, not time, so shot spacing was consistent. I also set the camera to program mode (not auto), set my (White Balance) WB to cloud or sunny (not auto) and ensure my shutter speed is above 1/1200. Camera is a SX260HS loaded with a CHDK script Jeff Taylor made – here’s a link: http://diydrones.com/profiles/blogs/apm-to-chdk-camera-link-tutorial” –Morgan

We’ve made some changes and improvements to the DroneMapper Aerial Imagery Processing service as of 02-21-2014.

• We no longer generate a non-blended geo-referenced Orthomosaic. (OrthoN-DroneMapper.tif)
• Orthomosaic GeoTIFF product has been renamed to Ortho-DroneMapper.tif
• 8 Bit DEM GeoTIFF product has been renamed to DEM8-DroneMapper.tif
• 32 Bit DEM GeoTIFF product has been renamed to DEM32-DroneMapper.tif
• DSM product has been renamed to DSM-DroneMapper.tif
• Low Resolution Point Cloud has been renamed to PointCloud_LR-DroneMapper.ply
• High Resolution Point Cloud has been renamed to PointCloud_HR-DroneMapper.ply
• Altitude Offset is no longer needed during imagery upload. Flight AGL is used instead.
• Geo-Referenced NDVI generation is an option during upload for NIR data collections.

Please contact us if you have any questions or comments!

We’ve added orthomosaic seamline detection and blending enhancement algorithms to the DroneMapper processing service. The routines have been tested on NIR imagery and RGB visible imagery with great results. These improvements significantly improve the overall visual appeal of our orthomosaic products without compromising the resulting geo-referenced NDVI or radiometrics. Please contact DN2K, AgEagle or DroneMapper for your precision agriculture needs. The NIR data set below is courtesy of DN2K and AgEagle.

DroneMapper has implemented blending and smoothing algorithms in order to eliminate or at least significantly minimize the appearance of seamlines in orthomosaics. Figure 1) shows a field imaged in the NIR with seamlines apparent throughout the field. A seamline detection algorithm was utilized to identify edges (shown in Figure 2) between processed tiles and apply a blending or feathering of the radiance at these seams. Figure 3 illustrates the result of incorporating this technique. The majority of seamlines have been significantly smoothed offering a more visually appealing scene. Figures 4 and 5 show the same area of the ortho at higher resolution before and after the technique was applied.

(Figure 1) Original Orthomosaic:

(Figure 2) Seamline detection:

(Figure 3) Orthomosaic with seamline detection and removal:

(Figure 4) Original Orthomosaic:

(Figure 5) Orthomosaic with seamline detection and removal:

Geo-Referenced Enhanced NDVI and Traditional NDVI: