For a school project my team and I are looking to detect nutrient defficiencies in plants through machine vision and deep learning. As gathering an extensive dataset of plants for an accurate enough detection is very labour intensive, we are looking to artifically create a large dataset by creating a 3D model of Bell Pepper plants and render these in Blender. In a previous project another student used a different modeling program to model the plant as this program (GROimp) offered the ability to easilly add variables to the model as to create some variety in the dataset.
We have seen that modeling plants in Blender is also possible, but would it also be possible to easily change the variables in order to not have to model a 1000 slightly different bell peppers? If this is the case, we can limit the software we use to Blender.
Thanks in advance for the advice, if the question needs further clarification just ask.