Search

Recently, image based plant phenotyping is used extensively for plant trait estimation, plant accession classification and selection, and plant stress analysis. Plant accession classification and selection helps in identification of plants tolerant to local climatic conditions. Over the past few years, convolutional neural network (CNN) models were predominantly used for classification of plants and plant diseases from static plant images using spatial features. In this paper, we introduce three different methods namely 3-dimensional (3-D) CNN, CNN with convolutional long short-term memory (ConvLSTM) layers and vision transformer that use temporal information along with spatial information for plant accession classification. We have used publicly available Arabidopsis plant accession classification dataset to test the performance of these methods. Time series color images of four different plant accessions of Arabidopsis are given as input to the neural network model which in turn identifies plant accession. All the three methods outperform the existing methods available in the literature in terms of average accession classification accuracy. Vision transformer achieves highest classification accuracy of 98.59% at the cost of very large number of trainable parameters as compared to the other two methods. On the other hand, CNN-ConvLSTM achieves comparable accuracy of 97.97% with very less trainable parameters as compared to vision transformer. In future, these models can also be used to identify unknown plant accessions and predict plant growth signatures in different climatic conditions.

Recent advancements in the area of computer vision-based plant phenotyping are playing an important role, in determining the quantitative phenotypes of plants, and crop yield. Automatic segmentation of plants and its associated structures is the first and most important step in image based plant phenotyping. We design and implement convolutional neural network (CNN) based modified residual U-Net for semantic segmentation of plants from the background. We also use SegNet and U-Net architectures for comparison purpose. In this paper, residual U-Net, SegNet and U-Net models are tested on leaf segmentation challenge (LSC) dataset and fig dataset that are publicly available. LSC dataset consists of images of Arabidopsis and tobacco plants grown under controlled conditions whereas fig dataset includes top view images of fig plants captured in open-field conditions. We have used 8 evaluation metrics for analyzing and comparing the performance of residual U-Net, SegNet and U-Net architectures with the existing algorithms in the literature. Residual U-Net with 15.32 million trainable parameters, outperforms SegNet and other state-of-the-art methods whereas achieves comparable performance with respect to U-Net. Residual U-Net achieves dice coefficient of 0.9709 on LSC dataset and 0.9665 on fig dataset, respectively. The segmentation networks used in this paper can be used for other plant related applications such as plant trait estimation or in quantification of plant stress.