Precision Ag Case Study

Hyperspectral Imaging Fights Citrus Blight

How can hyperspectral imaging make a difference to precision agriculture? Consider this example of a US project to inspect orange groves, looking for citrus blight disease, which used a Headwall Photonics’ Hyperspec® VNIR hyperspectral solution with an OEM camera mounted on a UAV.

Citrus blight destroys the vitality of trees and can spread throughout the grove. One of the early signs of this disease is a byproduct secreted on the surface of the leaves. Inspection for this used to require a person climbing a ladder to inspect the top of each tree. More often, growers might not know of a problem until trees started dying.

With hyperspectral imaging, this can be seen 300-400 meters above the crops covering a large area quickly (mounted on UAVs), allowing quick action to eliminate the spread and minimise the destruction.

Had only multispectral sensing been employed, this level of detail would not have been available and the correct decisions would not have been taken. Crop loss due to disease, such as citrus blight, and inaction leads to the loss of millions of dollars and the loss of a whole season in many cases.

PAS is the exclusive distributor of Headwall’s range of industry-leading hyperspectral solutions, and customises them for the demands of specific applications.

What’s the difference between multispectral and hyperspectral?

Imagine we could safely view the world through the eyes of different creatures, enabling us to view infrared radiation, ultraviolet light and reflected electromagnetic energy*. Well, we can, with the use of multispectral and hyperspectral sensors.

Each technology is able to sense (“see”) outside the range of normal human vision. The difference is the number of bands and the narrowness of the bands.

Multispectral imagery generally refers to 3-10 wide bands, using a remote sensing radiometer.

Hyperspectral imagery consists of many more bands (hundreds or thousands) that are much narrower (10-20nm), using an imaging spectrometer.

 

Why does it matter for precision ag?

It’s a question of detail.

The more detailed the spectral information recorded by a sensor, the more information that can be extracted from the spectral signatures.

Hyperspectral sensors have much more detailed signatures than multispectral sensors and thus provide the ability to detect more subtle differences in aquatic and terrestrial features.

Consider this example about a US project to inspect orange groves, looking for citrus blight disease, which used a Headwall Photonics’ Hyperspec® VNIR hyperspectral solution with an OEM camera mounted on a UAV.

Citrus blight destroys the vitality of trees and can spread throughout the grove. One of the early signs of this disease is a byproduct secreted on the surface of the leaves. Inspection for this used to require a person climbing a ladder to inspect the top of each tree. More often, growers might not know of a problem until trees started dying.  With hyperspectral imaging, this can be seen 300-400 meters above the crops covering a large area quickly (mounted on UAVs), allowing quick action to eliminate the spread and minimise the destruction.

 

For precision ag, the more detail, the better.

Imagine flying over a property and looking down at fields of different colours and appearance. You might decide you know what is planted in each paddock, with accuracy dependent on how low you are flying.

With multispectral detection, you would certainly be able to see differences between various elements, such as a tree plantation versus another crop.

With hyperspectral capabilities, you will see individual trees and plants, and even the subtle differences in the EMR emitted by disease and soil moisture levels etc.

In its 2017-2019 Strategic Plan, the NSW Department of Primary Industries’ first goal is Innovation in primary industries to improve resilience and boost productivity.

There is no doubt that hyperspectral imagery is the most advanced precision ag solution to transform our analysis and optimisation of production, risk management, use of resources and reducing disease.

 

Talk to PAS about the possibilities for the use of hyperspectral solutions in your agricultural context.

 

1. https://gisgeography.com/multispectral-vs-hyperspectral-imagery-explained/
2. https://www.adimec.com/how-is-hyperspectral-imaging-advancing-agriculture/

Access sophisticated precision ag technology for accurate and innovative farm management

Precision ag is a critical remote sensing application with the potential to impact any aspect of society. The use of UAVs in agriculture has been well-documented but the value of high-end data collectors mounted on the UAV or other vehicle is transforming data into agricultural treasure!

Peter Drucker’s saying, ‘what gets measured, gets improved’, highlights the importance of robust analysis and interpretation of data.

With precision ag, high-value crops can be planted, cared for, and harvested with the help of spectral data that leads to better decision-making. As a result, spectacular crop yields and healthier foods are achievable outcomes – of enormous value where agricultural success is a life-saving or life-enriching necessity.

Hyperspectral image sensors also monitor irrigation levels, pesticide and fertiliser effectiveness, and spot the telltale signs of invasive and hard-to-detect diseases. The key is to recognise these signs early and often, which is why continual airborne monitoring allows for trend analysis throughout the season.

From UAVs and aircraft as well as tractors and other mobile machinery, researchers can determine stress levels and overall plant vitality using hyperspectral image sensors at the VNIR (400-1000nm) and SWIR (900-2500nm) spectral ranges. These sensors operate in a line-scanning fashion, requiring movement to occur as the sensor builds a data-rich image cube containing all the spatial and spectral information within the field of view. GPS and LiDAR, plus the post-processing task of orthorectification, stamp the image data precisely. The end result is high-quality data that farmers and agriculturalists can use to make smart decisions.

 

Precision ag in Australia / New Zealand

In the Australian agricultural setting, hyperspectral imaging’s ability to cover distances with accuracy is a boon to primary producers seeking to maximise yield, minimise waste and manage scarce resources.

PAS is the Australian / New Zealand home of a range of sensors and analysers for any aspect of precision ag, including the Nano-Hyperspec and Micro-Hyperspec sensors from Headwall Photonics and the FieldSpec NIR models from ASD.

Headwall supplements its sensors with the industry’s best airborne hyperspectral software package called Hyperspec III. The package includes all the tools necessary to set up the sensor for airborne operation, and tie that operation in with GPS and LiDAR data streams. The software also manages more than one sensor at a time, permitting a single pass using a VNIR and a separate SWIR sensor for example.

Headwall’s sensors feature aberration-correction for precise image data from edge to edge. This wide field of view is particularly beneficial for the new breed of small, lightweight UAVs that require flight-path optimisation. The wider the field of view, the fewer passes over a plot of land the UAV (or aircraft) needs to make.

The ASD FieldSpec® line offers a wide range of configuration options for both contact measurements (such as leaves or in a soil profile pit) and stand-off measurements (such as those needed to measure canopy reflectance). The FieldSpec uses a flexible fibre optic cable that can be used with many different accessories and configurations, giving researchers many options for acquiring critical data.

Speak to PAS about the breadth of solutions available to meet the needs of precision ag applications and many others as well.