Titulo Estágio
Detection of agricultural pests in natural noisy images
Áreas de especialidade
Sistemas Inteligentes
Engenharia de Software
Local do Estágio
CISUC
Enquadramento
Agriculture has been evolving in a remarkable way since the agricultural revolution around 10,000 years ago, being increasingly technological together with the presence of great impact intelligent methodologies with applications in different areas such as the prediction of diseases and pests, the forecast of the best phase for harvesting or forecasting its volume.
However, agricultural technology is under immense challenges due to diseases and pests spread by globalization and aggravated by climate change. Portugal is one of the European countries that has been experiencing the impact of climate change, namely the increase in temperature, which increases the arrival and installation of new pests. Integrated protection is an area where the application of technologically advanced solutions is of evident interest.
In this internship, we propose the followup of a previous work on pest detection in images. While the first approach was to detect pests in sticky traps, it is much more interesting to detect pests in natural images, taken from the plants and fields (leaves, ground, etc.).
This process will entail research and implementation of deep learning models that allow detecting and locating pests both in plants and in traps using open source platforms like Google and other competitors, e.g. Tensor Flow, Keras that, when applied to images harvested in an agricultural environment, can be an important aid in the early detection of pests, thus contributing to their less economic and environmental impact.
Objetivo
In this internship the student should study, propose, implement, and test models for intelligent detection of plant pests in images.
To achieve this goal, the following objectives will be pursued:
- Study the state of the art
- Study the available datasets of natural images for pest detection
- Study the available frameworks for model development
- Define, implement, and fine tune the learning architecture
- Propose and deploy test setup
Plano de Trabalhos - Semestre 1
- Literature review
- Identification and annotation of datasets for model development
- Identification and study of available frameworks for model development
- Define the architecture of the system
- Start implementing the proposed approach
- Write intermediate report
Plano de Trabalhos - Semestre 2
- Implement the proposed solution and fine tune models
- Test and evaluate performance
- Write scientific article
- Write final report
Condições
This work should take place in the context of a research project in a CISUC lab.
This work should take place in the context of a research project funded by FCT in a CISUC lab. There is the possibility of a 6-month scholarship of 798 euros per month.
Observações
During the application phase, doubts related to this proposal, namely regarding the objectives and conditions, should be clarified with the advisors, by e-mail (catarina@dei.uc.pt) or a meeting to be arranged after a contact by email.
Orientador
Catarina Silva, Joana Costa, Bernardete Ribeiro
catarina@dei.uc.pt 📩