Top 3 Benefits of Hyperspectral Imaging for Agriculture Research

Traditional imaging recreates a scene, outputting spatial data as a 2D array of colour pixels. Spectroscopy measures a spectrum, outputting spectral data as a plot of wavelength against light intensity. 

Next-generation hyperspectral imaging cameras, including more robust and portable systems capable of video-rate read-out, are creating new research opportunities across a range of industries, perhaps the most active area of research is agriculture, where technological advances are pushing hyperspectral imaging out of the lab and into the field. 

High Spectral and Spatial Resolution in Agriculture Research 

Organic matter, such as plant tissue and soil, generates spectral signatures that reveal information about its structure and composition. While indetectable to conventional RGB devices, observing these signatures with hyperspectral imaging cameras reveals vital information for the agricultural industry. 

In fact, spectral imaging has been used for decades to enhance agricultural and environmental monitoring datasets. To date, however, hyperspectral cameras were typically employed as part of aerial remote sensing platforms. While these flight instruments provide valuable spectral data on soil conditions, vegetation health, and other big-picture trends, they are simply too far away to deliver plant-by-plant analysis. 

Now, new leaps in hyperspectral performance are pulling the technology down to earth, enabling in-situ, real-time assessment using ground-based sensing platforms and handheld measurements. High spectral resolution from hyperspectral imaging technology – combined with high spatial resolution from getting up close to the crops – could form the basis of new, more advanced agricultural practices that maximise output. 

Why the Agricultural Industry Needs Innovation 

The 2023 State of Food Security and Nutrition in the World report found the prevalence of undernourishment has grown significantly since pre-pandemic levels. Estimates put the number of people facing hunger in 2022 at between 691 and 783 million people, an increase of roughly 122 million since 2019. 

The report also discusses the need for new urban and peri-urban agriculture (UPA) technologies, due to continued urbanisation affecting agri-food systems. Stating: 

“The development of UPA is closely linked to the adoption of productive technologies and innovations, which can lead to increased yields and reduced environmental impacts. Considering the scarcity in urban areas of natural resources such as land and water needed for the production of nutritious foods, technology could play an essential role in making urban agriculture a sustainable alternative for food supply.” 

Researchers are investigating methods of incorporating spectral imaging data into agricultural processes to help maximise the food output for a given input of land and resources.  

This includes general crop monitoring techniques and more in-depth processes like constant real-time hyperspectral data capture to improve controlled environment agriculture (CEA) practices.  

The three most significant benefits of hyperspectral imaging for research in agriculture are: 

1. Identifying Trends Before They Are Visible 

With hyperspectral cameras, farmers can learn more about their crops and discover issues before they impact the entire harvest. Improved automation in quantification of crop health metrics is enabled by hyperspectral imaging. This is key to moving beyond spot sampling and delivering whole field monitoring with integration onto autonomous platforms. 

Real-time hyperspectral monitoring could enable early intervention, spotting stress factors from high spectral resolution data, rather than waiting for them to become apparent during visual inspection. A single hyperspectral system could provide early detection systems for many stress factors, such as: 

• Drought 

• Pests 

• Disease 

• Nutrient deficiency 

• Weeds 

Beyond identifying factors hindering growth, hyperspectral technology could also provide data for crop yield predictions, food quality assessment and shelf life estimates. Armed with this data, farmers would be better able to plan for the future, and businesses further down the supply chain could better predict the shelf life of products and reduce food waste. 

2. Use in the Field 

Hyperspectral hardware is becoming more compact, robust, versatile, and easier to use. This allows researchers to collect hyperspectral data where it makes the most difference – in the field, rather than in the lab. 

Systems based on line-scanned hyperspectral technology required more complex mechanical arrangements. They also introduced motion to pan across a scene, taking many frames to build up the hypercube. This leads to slower outputs while also being more fragile and complicated to use. 

By removing the need for motion and reducing complexity, advances in snapshot hyperspectral imaging have made it possible to build portable, handheld systems. Snapshot hyperspectral use cases are no longer limited by spectral resolution, and new cameras have doubled down on their main advantage – speed. These cameras can output video rate hyperspectral data for live assessments, while also offering a range of practical benefits in terms of ease of set-up and use.  

The goal is to transfer hyperspectral technology from sterile research laboratories and into the outside world. The challenge remains to abstract away low-level details on optical elements and compression techniques such that farmers around the world can use these systems to understand when to increase irrigation or apply fertiliser for maximum yield. More research is needed to design real-time algorithms that can keep pace with changing environments and use cases to support global food security. 

3. Integrating With Other Technologies 

Hyperspectral imaging can be integrated with a wide range of hardware and software technologies to improve agriculture research. This includes embedded machine vision cameras and IoT devices to provide the connectivity needed to automate agricultural processes. 

It also includes feeding hyperspectral data into cutting-edge machine learning algorithms for enhanced image processing. Hyperspectral cameras are the initial data collection stage of a broader technology stack. The raw hyperspectral dataset, providing pixel-level spectral information, can augment many computer vision applications in agriculture, and enable data-driven decision-making based on much better data. 

The Living Optics Camera 

The Living Optics camera, using next-generation snapshot technology, represents a leap forward for hyperspectral imaging. A portable, handheld, high-spectral resolution camera and real-time video software development platform; the Living Optics camera is a crucial enabling technology to advance agriculture research.

Contact the Living Optics sales team to learn more about the potential research benefits of the hyperspectral imaging camera.