Our first Successful NDVI Analysis Test

Good news. After months of painstaking builds, troubleshooting and configuration, we were finally able to get airborne for our first NDVI test!

The test involved a simple mission flying over a small field, and capturing images sequentially using a special NDVI camera attached to the drone.

IMG_20180601_161732
Our test equipment for the day

The camera itself is a 12-Megapixel image sensor that captures Red, Green and Near-Infrared (R-G-NIR) wavelengths in both JPG and RAW formats, ideal for post-processing analyses including Normalized Difference Vegetation Index (NDVI) calculations.

IMG_20180601_161903
NDVI Camera attached to front of drone, with cables for GPS and remote trigger functionality.

Covering only 0.15 hectares, the test took a total of less than 2 minutes, but that was more than enough time to gather vital data on the field for later processing. The field consists mainly of a dry, dusty section in the middle, and grass growing all around the perimeter.

IMG_20180601_161403
Taking flight to commence the test

Upon completion of the flight, the first order of business was to transfer all image files onto the computer and perform initial processing of the RAW data. The camera manufacturer supplies a software application that allows conversion of the RAW images to TIFF format, a more mainstream format that can then be used to develop precise maps with other third-party applications.

The first dataset to be developed in post-processing is known as a point cloud, which is a single large image developed from stitching multiple overlapping and georeferenced images together, to form an accurate, to-scale representation of a survey area.

101_Point Cloud
Point cloud map, showing the relative elevation and position of where the image was captured

Because TIFF format saves individual subpixel data (ie. Red, Green and NIR intensity), the mapping software can then develop an image based on indices such as NDVI ({NIR-Red}/{NIR+Red}), whose primary function is to analyse crop health and density. In this case, it was useful in analyzing the areas of the field that were dense with grass and those that were dry.

103_NDVI
NDVI image, showing the dry centre (red) and grassy perimeter (green) of the field

Further analyses can then be done using the above information, such as this thermal reflectance map, which would not only be useful in crop analysis, but also in analyzing solar panel and roof tile damage.

102_thermal reflectance
Thermal reflectance analysis. Healthy vegetation reflects high amounts of Infrared radiation.

To conclude, this was truly an informative and fulfilling experience. Watch this space as we seek to run pilot tests on farm plantations on our subsequent missions.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s