The presence of grass seed infestations in livestock poses significant biosecurity and economic challenges to the Australian agricultural industry. Traditional detection methods rely on manual inspection and histopathological analysis, which can be time-consuming, labour-intensive, and prone to human error. This project aims to leverage terahertz (THz) imaging technology to develop a non-invasive, rapid, and highly sensitive detection method for identifying infested grass seeds in sheep wool and skin. We will use a state-of-the-art THz camera from INO to capture high-resolution imaging data. By utilizing the unique spectral signatures of biological tissues and foreign contaminants in the THz range, this project seeks to establish proof-of-concept imaging techniques that enhance accuracy and efficiency in agricultural diagnostics.
School
Electrical Engineering and Telecommunications
Research Area
Terahertz technology | Imaging | Optics | Signal processing | Electrical engineering | Electromagnetics
- Research environment
- Expected outcomes
- Supervisory team
- Reference material/links
This project requires no prior knowledge of crucial system parameters, such as the specific THz spectral characteristics of different biological tissues. The primary focus will be on learning and extracting relevant features from THz imaging data to develop an efficient detection framework. The data acquisition process using the INO THz camera will be combined with image processing techniques to identify distinguishing patterns between infested and non-infested samples.
A research assistant, such as an undergraduate or postgraduate student, will be involved in the project under supervision. The student should have basic experience in MATLAB or Python for data processing and image analysis, along with an understanding of electromagnetic wave interactions with biological materials. Training sessions will be provided to familiarise the student with THz imaging principles and data interpretation methodologies. The project will be structured over two months, with an initial data collection phase, followed by algorithm development, testing, and reporting.
The research findings will contribute to advancing non-invasive agricultural diagnostics, with potential applications extending to biosecurity monitoring and precision farming techniques.
- Proof-of-Concept THz Imaging System: Develop and validate an initial THz imaging setup optimized for detecting grass seed infestations in biological samples.
- Preliminary Algorithm Development: Implement an initial classification algorithm to automate infestation detection based on THz spectral data.
- Comparative Analysis: Benchmark THz imaging performance against conventional detection methods (raster scanning) to assess improvements in speed, accuracy, and feasibility.
- Feasibility Assessment: Evaluate the practical applications and scalability of THz imaging for field deployment in agricultural settings.
- Final Report & Recommendations: Document findings, challenges, and recommendations for further research or potential commercialization.
- Thigale, S., Wang, Q., Mishra, D., Goldys, E. M., & Atakaramians, S. (2023). Terahertz imaging: a diagnostic technology for prevention of grass seed infestation. Optics Express, 31(22), 37030-37039.
https://opg.optica.org/oe/fulltext.cfm?uri=oe-31-22-37030&id=540968 - Wang, Q., Goay, A. C. Y., Mishra, D., Goldys, E. M., & Atakaramians, S. (2025). Diagnosing Grass Seed Infestation: Convolutional Neural Network Based Terahertz Imaging. IEEE Access.
https://ieeexplore.ieee.org/document/10843713 - Lee, W. S., Ferrante, A., Withayachumnankul, W., & Able, J. A. (2020). Assessing frost damage in barley using terahertz imaging. Optics Express, 28(21), 30644-30655.
https://opg.optica.org/oe/fulltext.cfm?uri=oe-28-21-30644&id=440158