University of Hong Kong Featured PhD Programmes
Birkbeck, University of London Featured PhD Programmes
University of Reading Featured PhD Programmes

Characterising individual Páramo plants and their health from UAV imagery using deep learning techniques


UK CEH

, , , Monday, January 04, 2021 Competition Funded PhD Project (Students Worldwide)

About the Project

The identification of plant individuals, their health and their dynamics is essential for environmental monitoring and management, whether this is in the context of plant population, habitat condition or ecosystem services assessment, invasive species or weed control action, or climate or disturbance impact assessment. While, currently, the most common approach is intensive field survey, the combined developments in unmanned aerial vehicles (UAV) and machine learning has opened up opportunities for automation.

This PhD project focuses on an iconic Páramo plant called frailejones. This plant is found in the Páramos, a highly diverse mountain ecosystem that is unique to the Northern Andes. Recently, Frailejones have been affected by moth caterpillars, beetle larvae and fungi, and have started to die in large numbers. It is unclear why this is happening. Some believe it is due to an increase in average temperatures, others blame a combination of environmental pressures. It is also unclear how many of the frailejones are affected and where, as their habitat, which lies above 4000 m altitude, is not easily accessible and covers a total area of 35,700 km2. Finding an effective approach to map and monitor the condition of individual plants is key to help evaluate current hypotheses. Frailejones typically have a single stemmed rosette of large leaves, that gives them the appearance of palms and both live and dead plants are generally easy to spot in the landscape on drone imagery and 3D point clouds. So automatically mapping them from drone data could be the solution. We have found that traditional image processing approaches are unsuitable, while artificial intelligence has shown great promise. This PhD will develop a novel deep learning framework for mapping live and dead frailejones using drone data. This deep learning framework will then be extended to include temperature, precipitation and land use as covariates to explore if these could be possible drivers behind the frailejones’ rapid decline

Through a currently NERC/AHRC funded project (PARAGUAS, https://paraguas.ceh.ac.uk), the student will have access to cm resolution UAV imagery, 3D point clouds, in conjunction with climate and land use data for a Páramo area in Colombia. This will provide context to the work and create opportunities for cross-disciplinary collaboration and a secondment of their choice. These experiences will be invaluable for the student’s development into a future environmental expert and leader.

The student will be based at the Centre of Excellence in Environmental Data Science (CEEDS; https://ceeds.ac.uk/), joint venture between UK Centre for Ecology & Hydrology (UKCEH) and Lancaster University, with collaboration in Imperial College London, Department of Life Science. Successful student will develop proficiency and pursue research on the following areas:
(i) Machine learning and deep learning techniques for image feature extraction.
(ii) Processing and analysis of ultra-fine resolution UAV imagery and 3D point clouds.
(iii) Development of data processing chains to map the health of species and to deliver products (plant distribution map) fit for long term monitoring.
(iv) Statistical analysis and ecological modelling to explore the potential drivers in the decline of endangered species.

Funding Notes

Applicants are expected to hold a first class or upper second class honours degree, and preferably also a Master’s degree or equivalent research experience, in a machine learning-related subject. You should have a strong background in computer science, mathematics, geography or environmental science with strong quantitative (e.g., programming) skills. You must have the demonstrable potential for innovative, high-quality PhD research. There are no residency restrictions for the SSCP DTP studentship, although there is more funding available for UK studentships, which make the International studentships highly competitive.

Please visit the SSCP DTP website for further funding and eligibility information (View Website).

References

Zhang et al., 2020, Identifying and mapping individual plants in a highly diverse high-elevation ecosystem using UAV imagery and deep learning. ISPRS Journal of Photogrammetry and Remote Sensing, 169: 280-291.
Zhang et al., 2020, Scale Sequence Joint Deep Learning (SS-JDL) for land cover and land use classification. Remote Sensing of Environment, 237: 111593.
Zhang et al., 2019, Joint Deep Learning for land cover and land use classification. Remote Sensing of Environment, 221: 173-187.
Zhang et al., 2018, A novel object-based convolutional neural networks (OCNN) for urban land use classification. Remote Sensing of Environment, 216: 57-70.

Email Now

Insert previous message below for editing? 
You haven’t included a message. Providing a specific message means universities will take your enquiry more seriously and helps them provide the information you need.
Why not add a message here

The information you submit to UK Centre for Ecology and Hydrology - Lancaster will only be used by them or their data partners to deal with your enquiry, according to their privacy notice. For more information on how we use and store your data, please read our privacy statement.

* required field

Your enquiry has been emailed successfully





FindAPhD. Copyright 2005-2020
All rights reserved.