Plant Image Analysis

Summary

Modern precision agriculture takes advantage of image data collected from various platforms to monitor crops and optimize their management and maximize yield. Plant productivity is significantly affected by various types of plant stress, such as those due to nitrogen, water, weed, or disease, which can be inferred from monitoring various reflectance indices. However, the classification of plant stress/severity levels especially based on limited data involving multiple stress types remains a major challenge. Similarly, automated plant data acquisition platforms, or even citizen science applications, require the ability to detect and recognize plant species in the wild. Despite a remarkable progress, this is still a challenging problem due to the variety of plant species, the unconstrained imaging conditions, and the variable reproductive state of plants, which also lead to a wide intra-species variability compared to the inter-species separation. This makes it a so-called fine-grained recognition problem (Keaton et al., 2021).

References

1. CVPRW

   Fine-Grained Visual Classification of Plant Species In The Wild: Object Detection as A Reinforced Means of Attention Keaton, M. R., Zaveri, R. J., Kovur, M., Henderson, C., Adjeroh, D. A., and Doretto, G. In Proceedings of the IEEE CVPR Workshop on Fine-Grained Visual Categorization, 2021. abstract bibTeX arXiv pdf data

   Plant species identification in the wild is a difficult problem in part due to the high variability of the input data, but also because of complications induced by the long-tail effects of the datasets distribution. Inspired by the most recent fine-grained visual classification approaches which are based on attention to mitigate the effects of data variability, we explore the idea of using object detection as a form of attention. We introduce a bottom-up approach based on detecting plant organs and fusing the predictions of a variable number of organ-based species classifiers. We also curate a new dataset with a long-tail distribution for evaluating plant organ detection and organ-based species identification, which is publicly available.

   @inproceedings{keatonZKHAD21cvprw,
     abbr = {CVPRW},
     author = {Keaton, M. R. and Zaveri, R. J. and Kovur, M. and Henderson, C. and Adjeroh, D. A. and Doretto, G.},
     title = {{Fine-Grained Visual Classification of Plant Species In The Wild: Object Detection as A Reinforced Means of Attention}},
     booktitle = {Proceedings of the IEEE CVPR Workshop on Fine-Grained Visual Categorization},
     year = {2021},
     arxiv = {2106.02141},
     data = {https://github.com/wvuvl/DARMA},
     bib2html_pubtype = {Conferences}
   }