Our mapping algorithms that do not rely on traditional image stitching techniques and high overlap are a key enabling technology allowing the system to fly at 20% overlap and produce maps offline, in minutes, on standard consumer laptops. This is the capability of SlantView: 30cm resolution maps (much higher than needed for creating even the most detailed variable rate prescriptions) generated from uncompressed native resolution data in the Map Window, combined with all the native resolution imagery in the Image Window.
Currently, the industry standard to generate native resolution stitched mosaics of highly aliasing-prone agricultural image data is a minimum of 80% overlap, resulting in 4x the flight time and 16x the raw data compared to SLANTRANGE systems. This data collection is often followed by mandatory cloud processing, requiring several hours of upload time even for users with the fastest internet service. Then, cloud processing on computers and servers much more powerful than anything readily available to consumers still takes hours. Software such as Pix4D allows local processing on high-powered computers with sufficient RAM, but takes hours to days, depending on the size of the data collection and available computing resources.
In many precision agricultural workflows, centimeter resolution data is critical. However, pretty pictures from centimeter resolution stitched mosaics are not practical given the corresponding raw data volume, processing power, and time requirements. Even the most detailed variable rate prescription and scouting applications do not require centimeter resolution maps over hundreds of acres.
If native resolution fully stitched mosaics are needed for your workflow, all SlantView licenses include the free Lite license for exporting your data for processing in Pix4D, Drone Deploy, and other native resolution mosaics stitching tools.
Selecting an Appropriate Map Resolution in SlantView
Select your map resolution based on the spacing of crops in the field. As a general rule, pick a resolution that is larger than the width of the empty space between plants. For example, if your crops are spaced 1 meter apart and 50cm resolution is selected, the colormaps shown in the Map Window may have "tiles" where no pixels of plant content exist, resulting in a map with blank areas between rows of crops. If your crop spacing is 30cm and 1m map resolution is selected, one "tile" of the Map Window can include content measured from several plants, and the color of that tile in the Stress Map will be a more accurate representation of the sum of the plant stress in that 1x1m area.
This is especially important when processing plant population data. The plant population maps shown in SlantView are area measurements. If the plant detections are oversampled, using a map resolution that is smaller than the spacing between plants, the population map shown in the SlantView Map Window will not present the data in a meaningful way. This is why the 30cm map resolution option is not available for Population data as it is for Stress analysis data in SlantView's processing settings menu.
It is important to note that SlantView is filtering out data from the soil, shadows, specular reflections, and other non-vegetation content. SlantView is not creating RGB maps, it is measuring reflectance data and displaying spectral ratios and crop health measurements in map form. Much of the edge detail recognized by the human eye does not translate to these NDVI, Stress, Vegetation fraction, and other SlantView map products that are visualizations of only the vegetation content in a given sample region.
Comments
0 comments
Please sign in to leave a comment.