Machine Vision for precision agriculture
Pneumatic precision vegetable planters are agricultural implements used to deposit vegetable seeds, such as carrots, onions, rucola, etc. at a prescribed distance on the ground.
The accuracy of the distribution and, consequently, the yield of the process, are directly linked on how well the seed distribution unit settings, which are manually set by the operator, based on her experience.
The goal of the thesis is to develop an advanced camera sensor, which uses machine vision algorithm to detect the quality of seed distribution and help the operator tasks.
The algorithm shall run on an embedded controller hardware, hence optimisation of algorithms to adapt to a limited resource and performance processors shall be considered.
Real-life test cases and all necessary mechanical parts and target hardware will be available to acquire data, test and validate the solution throughout the project duration.
The solution shall also include the evaluation and the use of Machine Learning algorithms to solve the tasks.
Required knowledge
- Computer vision algorithms
- Basic knowledge of Python/microPython programming languages
- Edge computing
Target domain: all MSc engineering courses
Notes: the thesis shall be carried out at ROJ company offices