DroneDeploy

How to Process Datasets

Cloud processing that scales with your business.

DroneDeploy's Map Engine is here to scale your productivity by offloading the processing, analysis, and sharing of your imagery. This gives you the ability to easily make use of the power, speed, and simplicity of DroneDeploy, regardless of what type of drone or camera you're using.

Once your data is processed, you will be able to make use of all the additional DroneDeploy functionality (such as exporting your data, sharing maps, annotations, and analysis).

Browser Requirements

Google Chrome or Mozilla Firefox are recommended to upload images. Internet Explorer may not allow you to upload images, and Safari may lead to degraded performance.

Please upload one dataset at a time. Uploading multiple datasets from multiple tabs on the same computer may cause errors in our system. If your internet connection is slow, consider using Turbo Upload as outlined below.

Data requirements

In order to use Map Engine, your data must meet the following requirements:

  • All images should be in JPG format
  • All images must have latitude, longitude, and altitude in the GPS EXIF data
  • All images should be facing the area of interest
  • All images should have significant overlap (more than 60% side and frontlap, 75% for agricultural or homogeneous imagery)
  • There must be at least 10 images. The maximum number of images changes according to your DroneDeploy plan.

As with all maps, input images directly affect map quality. Please read our spotlight on improving map quality here.

Considerations and Suggestions

eBee RTK data

SenseFly eBee imagery with embedded RTK is also compatible with DroneDeploy on all Business, Premier, and Enterprise plans. Read more about our compatibility with eBee RTK here.

Post processing of images.

Unless you know your editing software will not affect the metadata of your images, use the images directly from your SD card when uploading to DroneDeploy.

Mac users: importing images to iPhoto creates two copies. The image set in "Pictures" has the embedded GPS data, but the images in iPhoto do NOT have GPS information in the EXIF data. Be sure to upload from Pictures in order to meet the data requirements.

Large datasets: Image Name Duplication

DJI automatically restarts image names after one thousand images. This means larger flights will be named as DJI_001 from DJI_999, and then back to DJI_001.

Images with the same name will be considered duplicates.

In order to avoid this, please rename your images prior to combining them into one folder for upload. Some suggested workflows to do so are here.

Windows
Mac

How does it work?

Overview

Using Map Engine is as easy as 1, 2, 3.

  1. Capture Imagery
  2. Upload Imagery
  3. Analyze Data

We do all the hard work of processing and uploading the results. In this walkthrough, we've assumed that you have done #1, will dig deeper into #2, and provide links to #3.

Getting started

If you have flown with the DroneDeploy app...

After your flight, your DroneDeploy.com dashboard will be updated with a blue Upload icon to the left of the name of your flight. Click on this button, then proceed to the steps below.

Upload button for flown DroneDeploy mission.

Upload button for flown DroneDeploy mission.

Uploading to the Upload button will generate a Flight Log.

For customers who fly offline, we recommend restarting the mobile app while connected the Internet in order to facilitate synching to your DroneDeploy.com dashboard.

What if my flight won't synch?

If your DroneDeploy mission is not showing the Upload button, please proceed to the next step.

For all other datasets...

For datasets not captured with the DroneDeploy mobile app, please click on the theUpload button at the top left of your dashboard. Note: DroneDeploy missions can also be uploaded through this button.

Steps for uploading

Example Dataset

If you do not have a dataset available, we recommend that you download the example dataset of a massive rock near the beautiful Goat Rock Beach in Sonoma, CA.

Selecting either of the two Upload buttons above will open a dialog explaining more about Map Engine.

Step 1. Select your dataset type.

Terrain vs Structure

You will notice that we have two options for processing - Terrain or Structure. Why the two options? Each is optimized for different data types:

Terrain - These are best suited for large tracks of land such as agricultural fields or site surveys.
Structures - This is optimized for buildings, statues, objects, and other very three-dimensional structures. Datasets are limited to 500 images.

Terrain Upload option

Terrain Upload option

Why are Structures maps limited to 500 images?

Structures mode emphasizes the elevation of the images in order to create a more detailed 3D model. This mode is limited to 500 images in order to prevent larger datasets from timing out and to give the dataset the best chance of being a successful map.

Structures upload option, limited to 500 images.

Structures upload option, limited to 500 images.

This example is a natural scene, so we've chosen Terrain processing.

Tip: You can also rename your map right here at the point of upload.

Step 2. Select your photos.

After clicking the Select Photos button, you will be able to select all the images you would like to be processed:

Screenshot of photos being selected

Screenshot of photos being selected

Then, you will now see where your camera locations are (green dots), and the current boundary we have selected for you based on the camera locations (the blue outline).

Because they are directly on the boundary, the gray geotags in this example would not be included in the map. Expanding the area is necessary in this case.

The green dots represent the camera locations, the bounding box represents the crop boundary for the map.

The green dots represent the camera locations, the bounding box represents the crop boundary for the map.

Notice that in the top left, you see more information about the map, including the area, size, number of images, and estimated time to process.

You will see a request to select the subject matter type and select images. You can also rename your map right here at the point of upload.

Step 3. Consider selecting Turbo Upload.

Turbo Upload

Intended for customers who have limited internet, Turbo Upload compresses the files within your dataset for optimized uploading.

Enabling Turbo upload means:

Your upload will be up to 4x faster
Your map will be about half the resolution compared to standard upload

Notice how the dataset below goes from 120 MB to 12 MB when Turbo Upload is enabled.

Turbo Upload can make a large difference in processing time- in the example below, the dataset goes from 6 hours of processing to about an hour of processing.

The time estimates will show the minimum expected time range, and the maximum expected time range, based on historical data. The variance in the times is because the time it takes to stitch a map really does depend on the subject matter (how easy it is to find unique features across all images).

Step 4. Confirm upload.
Once you are happy with your settings, click the Upload button in the lower right to start the upload process. Your images will be uploaded and the data processing will start automatically.

When the upload has successfully completed, you'll see the following dialogue:

If you now look at your dashboard, you'll see a new item has been added with the name you selected.

When we've completed the processing, you will be notified via email. Processing the high-quality data can take up to a few hours for a very large job with high-resolution imagery. On average, processing takes about one minute per image. If you've chosen "Structures" from the options above, processing may take longer to complete.

Step 5. Access map via email

When your map is complete, you will receive an email notification with further instructions.

Where to from here?

Congratulations on your first map! Learn more about the data presented in each may layer here. Be sure to check out how to Crop, Annotate and Share your maps!

How to Process Datasets

Cloud processing that scales with your business.