Applications of Hyperspectral Imaging in Agriculture

↖ Like, please follow/top, Agriculture New Perspectives, thank you

Hyperspectral imaging (HSI) has become a promising research tool in precision agriculture (PA). Unlike visible light (RGB) or multispectral imaging technologies, hyperspectral imaging technology provides a more detailed and comprehensive understanding of crop health, enabling more targeted and precise crop management decisions. HSI can detect plant stress, diseases, and nutrient deficiencies by analyzing the unique spectral characteristics of crops, providing valuable information to optimize crop yields and reduce input costs.

Initially, HSI was mainly used in orbital and suborbital platforms. Now, portable handheld versions of HSI have emerged and are widely used in engineering, medical scientific research, and industrial production lines. In the past two decades, the number of peer-reviewed studies using handheld hyperspectral sensors has been increasing, with more research building customized data collection platforms and analytical workflows (Figure 1).

Figure 1 The trend in peer-reviewed publications shows an increasing number of studies utilizing handheld hyperspectral sensors and customized platforms for hyperspectral data collection or analysis.

Although hyperspectral sensors have been continuously improved over the past few decades, they still lack the general feasibility for real-time applications compared to RGB cameras. Hyperspectral data has a large volume compared to digital images, with high data acquisition costs and workloads, and data analysis is mostly conducted in research laboratories, limiting its adaptability in the research community, making real-time utilization of hyperspectral data impractical. To overcome the challenges of hyperspectral sensors in real-time applications, we need to shift our focus to technological approaches to bridge the gaps that arise when processing high-dimensional hyperspectral data. Table 1 highlights key areas of recent review studies on the application of hyperspectral imaging in agriculture.

Recently, Billy G. Ram et al. published a review titled A systematic review of hyperspectral imaging in precision agriculture: Analysis of its current state and future prospects in Computers and Electronics in Agriculture, mainly introducing the current application status and future prospects of hyperspectral imaging (HSI) in precision agriculture. Billy G. Ram et al. conducted a systematic review of 163 scientific articles published over the past two decades (2003-2023). Among them, 97 articles were selected for further analysis based on their relevance to the current topic. The topics include traditional data preprocessing techniques, hyperspectral data acquisition, data compression methods, and segmentation methods. The hardware implementations for high-speed data processing and machine learning applications using field-programmable gate arrays (FPGA) and graphics processing units (GPU) were discussed.

This review focuses on the real-time application requirements of ground hyperspectral sensors, emphasizing integrated ground, drone, and airborne hyperspectral sensors with high spatial resolution. It highlights the enormous potential of ground hyperspectral imaging technology in precision agriculture and prospects as a research tool for real-time crop monitoring and analysis. It discusses its limitations and the hope that the development of HSI technology brings for future agriculture and food security.

Table 1 An overview of significant hyperspectral imaging research conducted in the field of agricultural engineering in recent years.

Figure 3 The literature screening process according to PRISMA guidelines.

Figure 4 Different modes of hyperspectral data acquisition.(a) Point scanning, (b) Line scanning, (c) Wavelength scanning and (d) Snapshot scanning.

Figure 5 Examples of hyperspectral platforms applied in the field.

(a) Autonomous Platform Information System (API); (b) Data collection with handheld cardboard background; (c) A portable spectrometer integrated from two spectrometers, operating on Raspberry Pi 3 for data collection (point scanning method); (d and i) Testing of platforms with micro-dose weeding functions under field (d) and laboratory (i) conditions; (e and f) Hypercart, an artificial lighting system for data collection; (e) and imaging box (f), examples of hyperspectral platforms for laboratory applications; (g) Integrated spectral imaging (ASI) for data collection and model deployment; (h) Coffee bean defect detection system using a robotic arm for sorting; (j) DJI M600 integrated NVIDIA Jetson TK1 data acquisition and recording system, using Specim FX10 designed for ground applications; (k) DJI M600 integrated drone-specific sensors with built-in data recording capabilities; (l) DJI S1000 integrated Rededge multispectral sensor.

Table 2 Various hyperspectral platforms used in agricultural applications.

Figure 6 (a) Visualization of hyperspectral data cube; (b) Hyperspectral images at different wavelengths; (c-d) Basic segmentation process using k-means clustering to remove background; (e) Display of original crop spectral characteristics; (f) Example of preprocessed spectral characteristics using Savitzky Golay second derivative preprocessing.

Figure 8 Various examples of data augmentation techniques based on spatial transformations applied to hyperspectral images. Red markers are used to highlight the changes in image orientation caused by these transformations.

Join our knowledge circle (more paid materials sharing):

For group communication, contact the editor, my surname is Yu, not Qian Xiaodong himself

WeChat ID: nongyeshuzihua123

Applications of Hyperspectral Imaging in Agriculture

RECOMMEND

Previous Recommendations

1、[New] Comprehensive Guide to the Agricultural Industry Chain (Join the group at the end)

2、Must-see for Agriculture (Latest Version) China Agricultural Map [Recommended for Collection]

3、[Agricultural Digitalization] Official Account Introduction + Historical Good Articles

4、These are the 8 Best Agricultural Majors in China

5、Data Sharing: China Crop Distribution Map

6、66-page PPT | Smart Agriculture Platform Solution

7Guidelines for the Declaration of Agricultural Projects Involving Technology Innovation and Technical Transformation Re-loan

8、[Fresh Industry] 28 PPT Analysis: Current Status of Fruit E-commerce Development and Shaanxi Practices

9、“4+1+N” Agricultural Industry Development Model

10、Is the Core of Developing Digital Agriculture Data or People

11、Fuxi Farm: Building a Smart Agriculture Data Foundation

12、What Crops Can Be Grown in High-Standard Farmland? (Including a Life Cycle Diagram of Crops)

Applications of Hyperspectral Imaging in Agriculture

Understanding the Relationship Between Micro:bit and Arduino

Getting Started with the Arduino ESP32 Core Library

Implementing Neural Networks on a 10-Cent RISC-V MCU Without a Multiplier

Getting Started with Arduino: Light Up Your First LED from Scratch

How Highly Integrated Embedded Processors Drive the Development of Industrial Robots

Leave a Comment Cancel reply

Menu