A stomata classification and detection system in microscope images of maize cultivars

Abstract

Plant stomata are essential structures (pores) that control the exchange of gases between plant leaves and the atmosphere, and also they influence plant adaptation to climate through photosynthesis and transpiration stream. Many works in literature aim for a better understanding of these structures and their role in the evolution process and the behavior of plants. Although stomata studies in dicots species have advanced considerably in the past years, even there is not much knowledge about the stomata of cereal grasses. Due to the high morphological variation of stomata traits intra- and inter-species, detecting and classifying stomata automatically becomes challenging. For this reason, in this work, we propose a new system for automatic stomata classification and detection in microscope images for maize cultivars based on transfer learning strategy of different deep convolution neural netwoks (DCNN). Our performed experiments show that our system achieves an approximated accuracy of 97.1% in identifying stomata regions using classifiers based on deep learning features, which figures out as a nearly perfect classification system. As the stomata are responsible for several plant functionalities, this work represents an important advance for maize research, providing an accurate system in replacing the current manual task of categorizing these pores on microscope images. Furthermore, this system can also be a reference for studies using images from different cereal grasses.

Morphophysiological and developmental parameters of maize varieties

“…These evaluated stomatal parameters act in an antagonistic way between them, that is, the higher the stomatal functionality, which implies a larger stomata, the lower the stomatal density, because smaller stomata will have a higher number per square millimeter (Castro et al, 2009). Thus, a doubt is created, in which future studies will have to answer, because some lines of research indicate that the stomatal density ensures greater efficiency in the photosynthetic rate of plants, and also a very large stomatal opening may favor the chemical and biological contamination of plants (Shrestha et al, 2021;Manjunatha et al, 2019;Aono et al, 2021. No statistical differences were observed among the leaf morphological variables of the maize varieties as presented in Table 5.…”

Foliar stomata characteristics of tree species in a university green open space

“…The stomata type was categorized into 4 (four) different types of stomata, namely anomocytic, anisocytic, paracytic, and diacytic (Da Silva et al 2014). To determine the density of stomata, the number of stomata was divided by the area of the observed cross-section (Aono et al 2021). The length and width of stomata were analyzed by using the same application based on the length.…”

“…The paracytic stomata are characterized by joining the guard cell with one or more of the adjacent neighboring cells with the longitudinal axis parallel with the axis of the neighbor cell and the aperture (Aono et al 2021). This type of stomata was found in 76 species, including Hura crepitans, Acacia mangium, Persea americana, Elaeocarpus grandiflorus, and Garcinia atroviridis.…”

No cell is an island: characterising the leaf epidermis using epidermalmorph, a new R package

“…However, as noted by Vőfély et al . ( 2019 ), accurate segmentation of images is nontrivial and poses a significant obstacle to high‐throughput studies (though see recent work on stomatal detection by Fetter et al ., 2019 ; Aono et al ., 2021 ; Xie et al ., 2021 ; Li et al ., 2022 ). Another obstacle to automation is the multitude of types of epidermal preparation and imaging methods resulting in many different types of input image (see Fig.…”

“…Automated tracing and annotation of cells to reduce processing time and errors is an exercise in semantic segmentation (see Marmanis et al, 2016), and it may be possible to train a machine learning algorithm to perform this step. However, as noted by V} of ely et al (2019), accurate segmentation of images is nontrivial and poses a significant obstacle to high-throughput studies (though see recent work on stomatal detection by Fetter et al, 2019;Aono et al, 2021;Xie et al, 2021;Li et al, 2022). Another obstacle to automation is the multitude of types of epidermal preparation and imaging methods resulting in many different types of input image (see Fig.…”