Abstract : Images play a vital role in crop phenotyping. Pixel-wise classification (into vegetation/background) or semantic segmentation is a critical step in the computation of several canopy state variables. Current state of the art methodologies based on convolutional neural networks are trained on data acquired under controlled environments. These models are unable to generalize to real-world dataset and hence need to be fine-tuned using new labels. This motivated us to create the P2S2 segmentation dataset – a collection of multi-crop RGB images from different acquisition conditions. We present here the dataset and state of the art results.
https://hal.inrae.fr/hal-03140124 Contributor : Marie WeissConnect in order to contact the contributor Submitted on : Friday, February 12, 2021 - 3:32:44 PM Last modification on : Monday, May 17, 2021 - 12:00:05 PM Long-term archiving on: : Friday, May 14, 2021 - 9:33:08 AM
Simon Madec, Kamran Irfan, Etienne David, Kaaviya Velumani, Gaetan Daubige, et al.. The P2S2 segmentation dataset: annotated in-field multi-crop RGB images acquired under various conditions. 7th International Workshop on Image Analysis Methods in the Plant Sciences (IAMPS), Jul 2019, Lyon, France. ⟨hal-03140124⟩