Automated analysis of foraminifera fossil records by image classification using a convolutional neural network

Abstract

Abstract. Manual identification of foraminiferal morphospecies or morphotypes under stereo microscopes is time consuming for micropalaeontologists and not possible for nonspecialists. Therefore, a long-term goal has been to automate this process to improve its efficiency and repeatability. Recent advances in computation hardware have seen deep convolutional neural networks emerge as the state-of-the-art technique for image-based automated classification. Here, we describe a method for classifying large foraminifera image sets using convolutional neural networks. Construction of the classifier is demonstrated on the publicly available Endless Forams image set with a best accuracy of approximately 90 %. A complete automatic analysis is performed for benthic species dated to the last deglacial period for a sediment core from the north-eastern Pacific and for planktonic species dated from the present until 180 000 years ago in a core from the western Pacific warm pool. The relative abundances from automatic counting based on more than 500 000 images compare favourably with manual counting, showing the same signal dynamics. Our workflow opens the way to automated palaeoceanographic reconstruction based on computer image analysis and is freely available for use.

An Improved Cleaning Protocol for Foraminiferal Calcite from Unconsolidated Core Sediments: HyPerCal—A New Practice for Micropaleontological and Paleoclimatic Proxies

“…The efficiency of the HyPerCal procedure in the cleaning of calcitic microfossils makes it complementary for foraminifera shell weight studies since it was shown to bring the measured weight closer to that of an "original" shell. Furthermore, it paves the way for its use in modern analytical techniques that require some degree of automatization, such as image recognition software that are unable to recognize a lot of foraminifera images, whose umbilical aperture is not fully cleaned and is infilled with remaining nannofossil ooze [28]. On the other hand, it has proved beneficial for the upcoming practice of microfossil X-ray tomography, since CT image analysis software cannot easily discriminate between contaminated areas and areas referring to the foraminifera tests unless (subjective) manual labor intensive segmentation is employed [24].…”

“…Residual clays or nano-ooze in poral spaces and shell surface obstruct the study of test ultrastructure that yield information about the degree of carbonate dissolution [25] or test porosity [26]. Furthermore, such coatings or infillings (in apertures) often precludes automated recognition software, which is based on morphological features of foraminifera shells [27], from classifying their images correctly [28] and greatly complicate specimen segmentation when using high resolution X-ray tomographic techniques [29]. In the present study, by using light microscope imaging, SEM and X-ray tomography to assess the cleanliness of tests treated with reagents that are established not to alter the fossil geochemical signal, we propose a methodology that effectively diminishes surface and internal specimen contamination.…”

An Improved Cleaning Protocol for Foraminiferal Calcite: HyPerCal – A New Practice for Micropaleontological and Paleoclimatic Proxies

“…The efficiency of the HyPerCal procedure in the cleaning of calcitic microfossils makes it complementary for foraminifera shell weight studies since it was shown to bring the measured weight closer to that of an "original" shell. Furthermore, it paves the way for its use in modern analytical techniques that require some degree of automatization, such as image recognition software that are unable to recognize a lot of foraminifera images, whose umbilical aperture is not fully cleaned and is infilled with remaining nannofossil ooze [28]. On the other hand, CT image analyses software cannot easily discriminate between contaminated areas and areas referring to the foraminifera shell unless (subjective) labor intensive segmentation is employed [24].…”

“…Residual clays or nano-ooze in poral spaces and shell surface obstruct the study of test ultrastructure that yield information about the degree of carbonate dissolution [25] or test porosity [26]. Furthermore, such coatings or infillings (in apertures) often precludes automated recognition software, which is based on morphological features of foraminifera shells [27], from classifying their images correctly [28] and greatly complicate specimen segmentation when using high resolution Xray tomographic techniques [29]. In the present study, by using SEM and X-ray tomography with reagents that are established not to alter the fossil geochemical signal, we propose a methodology that effectively diminishes surface and internal specimen contamination.…”

Semantic segmentation of vertebrate microfossils from computed tomography data using a deep learning approach

“…DL-based techniques have achieved excellent performance in various computer vision tasks such as image denoising (Tian et al, 2020), target detection (Khan et al, 2017), image classification (Xu et al, 2020), and image segmentation (Jin et al, 2018). Paleontologists are utilizing the capabilities of deep neural networks (DNNs) to solve paleontological problems (Marchant et al, 2020;Tetard et al, 2020;Bourel et al, 2020). DNNs can be exploited not only for the accurate classification of vertebrate fossils from their 3D volumes (Hou et al, 2020), but also for the rapid documentation of discrete fossiliferous levels (Martín-Perea et Semantic image segmentation is also an important application of deep learning that separates a single image into different parts.…”