Non-destructive inspection for plant stress detection, disease diagnosis and nutrient deficiency analysis

Hyperspectral Datasets Help Inspect Crops at Greater Scale and Speed, Identifying and Intervening Early to Minimise Exposure

Non-Invasive Identification and Early Detection of Plant Disease?

Hyperspectral imaging and SPAD measurement for plant health monitoring and disease detection using image processing techniques

The reliable and timely identification of plant diseases is a crucial challenge in modern agriculture. Early detection of conditions impacting the health of plants, like insects, fungus, or nutritional deficiency, can help researchers to prevent biotic and abiotic stresses in plants.
Conventional methods rely on manual observation of visible symptoms. Manual methods can be time-consuming, labour-intensive, and limited to the late stages of infection. Visible symptoms tend to exhibit at the middle or late stages of infection which increases the likelihood of spread or reduced yields. Traditional methods of chemical analysis, require destructive sampling of plant tissue which can be time consuming and further stress a diseased plant.

To address these limitations, there is a growing interest in utilizing automated and objective approaches for early disease identification. Image analysis techniques, particularly using hyperspectral and RGB imaging sensors, offer promising solutions for precise and non-invasive disease detection.

What are the methods of detection of plant disease?

There is a growing body of research in the field of image analysis techniques, machine learning algorithms, and high-throughput phenotyping for stress identification and prediction. By leveraging various imaging sensors, including RGB, hyperspectral, thermal, and chlorophyll fluorescence, these approaches provide objective and sensitive detection of biotic and abiotic stresses in plants.
The adoption of automated disease identification techniques, such as hyperspectral and RGB imaging combined with advanced image analysis and machine learning, holds great promise for improving crop management practices. These non-invasive and objective approaches enable early detection of plant diseases, facilitating timely intervention and precise resource allocation. By leveraging advancements in technology and data analysis, researchers are driving the development of more efficient and accurate disease identification methods, paving the way for sustainable agriculture and improved crop health.
Hyperspectral imaging (HSI), an emerging technology, captures detailed reflectance information across a broad spectrum of light that extends beyond human vision. This allows for the identification of subtle changes in plant growth and development, such as the accumulation of pigments like anthocyanin. When exposed to stress conditions like high-intensity light or nutrient deficiency, plant growth and yield are affected. The accumulation of anthocyanin is an early indicator of this stress. Usually accumulation starts in a small region, making it difficult to see.

By measuring the intensity of light reflected or absorbed by the plant, a HSI camera can identify changes within specific spectral ranges to pinpoint changes in the plant’s biological, chemical and physical properties. The Specim IQ is an innovative solution that brings simplicity and portability to HSI to provide entry level access to this potentially transformative technology that can be used both in a lab or in the field.

Specim IQ - Accessible and Versatile Hyperspectral Imaging

The Specim IQ is a handheld hyperspectral camera designed with camera-like and simplicity; Point at the target, define the measurement settings, record and view data.

Comprehensive hyperspectral camera system based on push-broom (line scan) technology, the Specim IQ has an easy-to-use graphic interface with classification and visualization tools. Additional modelling, reporting and analysis is possible via the Specim IQ Studio software.

In the field of plant pathology and monitoring crop health, plant specific datasets can be measured to identify areas of the electromagnetic spectrum where changes in reflected or absorbed spectra can be used to either identify disease or other plant health indicators. HSI can then be used to provide comparatively quick targeted investigations potentially far quicker and more regularly than would otherwise be possible. Results and insights can be achieved without the need for complex mathematics or extensive knowledge of HSI and without the need for destructive sampling.

Chlorophyl Meter SPAD-502 Plus

The SPAD meter is widely used within commercial agriculture for performing non destructive testing of crop health, providing simple yet powerful data that allows growers to calculate the optimum timing and quantity of fertiliser. SPAD readings correlate with chlorophyl present in the leaf and calculation of concentration.

Performing routine testing across the plot to monitor plant health and optimise fertiliser use can improve yields whilst reducing costs and reducing the environmental load.


  • Non destructive / Non invasive approach
  • Detect and identify plant disease and stress before visible symptoms
  • Intervene and monitor the disease to prevent its spread
  • Quantify the percent area of leaf effected by disease
  • Identification of stress symptoms in plants (presence and accumulation of anthocyanin) based on water content or photosynthetic status
  • Optimise timing and quantity of nitrogen fertiliser and nutrient conditions: reduce waste and cost of overuse (over-fertilization), thereby reducing environmental contamination (diseases in plants and water contamination due to nutrient leaching through the field’s soil)

Benefits of Automated Disease Identification:

Automated disease identification enables early intervention, targeted chemical application, and improved crop management practices. By detecting diseases at their early stages, farmers can prevent the spread of infection and reduce crop damage. Precise disease identification allows for targeted chemical usage, reducing pesticide and herbicide application, cost, and benefiting the environment.



SPECIM IQ Hyperspectral Camera

A handheld hyperspectral camera designed with camera-like and simple usability; Point at the target, define the measurement settings, record and view data.

SPAD-502 Plus

Simple, portable, non destructive testing of plant leaves to measure chlorophyl content, allowing a fast and easy route to optimise the timing and quantity of fertiliser. 



Use of a Specim IQ for investigation of pathogen detection and excess water content.

The Specim IQ was used to build data models for detection of root rot and plant health, identifying particular wavelengths of significance within different plant types.

Source: ResearchGate Alt et al. (2020)

Read More



Combining HSI with LSSVM (least squares support vector machine) to Identify Disease in Citrus Trees

Hyperspectral Imaging used for characterising identifying HLB infected citrus trees. Creating a dataset that helped to distinguish dehydrated trees from diseased trees so that spread of infections can be limited, leading to a reduction in costly losses and unnecessary removal of falsely diagnosed stock.

Source: ResearchGate Weng et al. (2018)

Read More

Early Disease Detection in Wheat Using HSI

Investigation into the characterisation of the spectral reflectance of diseased wheat plants in-field to assess the ability to combat yellow rust with targeted interventions and minimise environmental load and cost.

Source: Science Direct Bravo et al. (2003)

Read More