Abstract
Poaceae represent one of the largest plant families in the world. Many species are of great economic importance as food and forage plants while others represent important weeds in agriculture. Although a large number of studies currently address the question of how plants can be best recognized on images, there is a lack of studies evaluating specific approaches for uniform species groups considered difficult to identify because they lack obvious visual characteristics. Poaceae represent an example of such a species group, especially when they are non-flowering. Here we present the results from an experiment to automatically identify Poaceae species based on images depicting six well-defined perspectives. One perspective shows the inflorescence while the others show vegetative parts of the plant such as the collar region with the ligule, adaxial and abaxial side of the leaf and culm nodes. For each species we collected 80 observations, each representing a series of six images taken with a smartphone camera. We extract feature representations from the images using five different convolutional neural networks (CNN) trained on objects from different domains and classify them using four state-of-the art classification algorithms. We combine these perspectives via score level fusion. In order to evaluate the potential of identifying non-flowering Poaceae we separately compared perspective combinations either comprising inflorescences or not. We find that for a fusion of all six perspectives, using the best combination of feature extraction CNN and classifier, an accuracy of 96.1% can be achieved. Without the inflorescence, the overall accuracy is still as high as 90.3%. In all but one case the perspective conveying the most information about the species (excluding inflorescence) is the ligule in frontal view. Our results show that even species considered very difficult to identify can achieve high accuracies in automatic identification as long as images depicting suitable perspectives are available. We suggest that our approach could be transferred to other difficult-to-distinguish species groups in order to identify the most relevant perspectives.
Highlights
Automated species identification is becoming an important and widely used tool to monitor the occurrence of species across a wide taxonomic range (Durso et al, 2021; Høye et al, 2021; Joly et al, 2021; Mahecha et al, 2021)
The Top-1 accuracies for the individual perspectives averaged across all species range from 87.5 to 26%, 75.3 to 17.2%, 70.1 to 18.2%, 64.9 to 17.6%, 63.7 to 13.1%, and 62.3 to 17.6% (Figure 8)
The features derived from the Flora Incognita convolutional neural networks (CNN) combined with an Support vector machines (SVM) classifier always achieve the highest accuracies, while the Open Images features combined with the Naive Bayes classifier always achieve the lowest accuracies for all single perspectives
Summary
Automated species identification is becoming an important and widely used tool to monitor the occurrence of species across a wide taxonomic range (Durso et al, 2021; Høye et al, 2021; Joly et al, 2021; Mahecha et al, 2021). About 12,000 species and 780 genera of Poaceae are described (Christenhusz and Byng, 2016; Soreng et al, 2017) which ranks them among the most diverse plant families worldwide. With only a few exceptions all Poaceae species are characterized by a unique set of characters that allows an easy attribution of individuals as members of the this family (). This more or less uniform morphology leads to the common perception of “grass” as a single species in the public (Jäkel and Schaer, 2004; Thomas, 2019). The sometimes very subtle differences between species or even genera can only be recognized by careful examination, especially if no flowers are present
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
