Our goals for the near future
When we consider yield estimation of cotton,
there are some existing solutions that try to solve this problem but they
fail to achieve the required accuracy because of the challenges of occlusion
and accurate cotton-boll counting techniques.
We overcome these issues by
developing techniques that capture images with minimum occlusion,
stitch and then analyse them using machine learning models that are best
suited for their process.
Another domain we are looking at is the quantitative phenotyping which uses an android phone. We are trying to develop an user friendly app which can be done with not much of a sophisticated setup. This automation speeds up the process of phenotyping by almost 20x. This helps the breeders to take the decisions quicker.
