Automated stomata detection in oil palm with convolutional neural network

Abstract

Stomatal density is an important trait for breeding selection of drought tolerant oil palms; however, its measurement is extremely tedious. To accelerate this process, we developed an automated system. Leaf samples from 128 palms ranging from nursery (1 years old), juvenile (2–3 years old) and mature (> 10 years old) were collected to build an oil palm specific stomata detection model. Micrographs were split into tiles, then used to train a stomata object detection convolutional neural network model through transfer learning. The detection model was then tested on leaf samples acquired from three independent oil palm populations of young seedlings (A), juveniles (B) and productive adults (C). The detection accuracy, measured in precision and recall, was 98.00% and 99.50% for set A, 99.70% and 97.65% for set B, and 99.55% and 99.62% for set C, respectively. The detection model was cross-applied to another set of adult palms using stomata images taken with a different microscope and under different conditions (D), resulting in precision and recall accuracy of 99.72% and 96.88%, respectively. This indicates that the model built generalized well, in addition has high transferability. With the completion of this detection model, stomatal density measurement can be accelerated. This in turn will accelerate the breeding selection for drought tolerance.

Rapid non-destructive method to phenotype stomatal traits

“…If the images are not satisfactory after this long process, other leaf imprints need to be taken, which is also time consuming. Although some recent methods included a machine-learning program to analyse stomata automatically, they still require taking leaf imprints using nail polish and taking images using a light microscope (8,14,15).…”

“…To overcome this limitation, many methods have included a machine-learning program to accelerate image analysis. However, most of these programs were developed based on images obtained from nail polish imprints (13,14,15). Although machine-learning approach reduces the time taken to acquire data, obtaining satisfactory images using nail polish remains difficult and the method cannot be applied on a large scale.…”

“…Recently, a number of methods have been published which address the issues with manual processing by including a machine learning algorithms that automatically detect and analyse stomatal features (16)(17)(18)(19)(20). While this improvement significantly accelerated image analysis and offers an effective substitute to manual analysis, these methods are limited by providing either information on stomatal number, size or aperture in one species in isolation; having data on all stomatal traits is desirable to have a comprehensive view on stomatal phenotype.…”

RotatedStomataNet: a deep rotated object detection network for directional stomata phenotype analysis

“…To systematically obtain phenotypic data of stomata, an end-to-end stomatal detection and identification framework based on feature weight transfer learning and YOLOv4 (You Only Look Once version 4) network was developed , which can locate stomata based on horizontal detection boxes and obtain phenotypic data such as the number, length, and width of stomata. To accelerate the selection of drought-tolerant oil palm breeders, MobileNet was used as a template to automatically detect stomata in oil palm to obtain the density of stomata in drought-tolerant oil palm (Kwong et al 2021).…”

Automated and high throughput measurement of leaf stomatal traits in canola

“…Automation of stomatal trait measurement using machine learning-based model has already been reported for other industrially important crops such as rice, wheat, tomato (Pathoumthong et al, 2023), barley (Sai et al, 2023) and oil palm (Kwong et al, 2021). However, to date, none exist for canola.…”