captured from the side view. The total number of photomicrographs obtained for each mor- photype is shown in Table 1. Only nonarticulated (not attached to any other phytoliths) were considered and just one photomicro- graph per phytolith was recorded. The total number of samples was 429. All the images are publicly available at https://reposi-tori.upf.edu/handle/10230/44939, all the morphotypes have at least 50 samples, and the dataset is fairly balanced (i.e., there is a similar number of samples per class). The image of each phytolith was digitized, using an open- source web annotation tool called VGG Image Annotator.1 This tool allows a researcher to create a control-points based contour
Open resource ↗reposi-tori.upf.edu · 10230/44939 · pdf-raw-page:3 lines:1-92Unverified paper record
"You Are Not My Type": An Evaluation of Classification Methods for Automatic Phytolith Identification.
Microscopy and microanalysis : the official journal of Microscopy Society of America, Microbeam Analysis Society, Microscopical Society of Canada · 1 Dec 2020 · 10.1017/s1431927620024629
Abstract
Phytoliths can be an important source of information related to environmental and climatic change, as well as to ancient plant use by humans, particularly within the disciplines of paleoecology and archaeology. Currently, phytolith identification and categorization is performed manually by researchers, a time-consuming task liable to misclassifications. The automated classification of phytoliths would allow the standardization of identification processes, avoiding possible biases related to the classification capability of researchers. This paper presents a comparative analysis of six classification methods, using digitized microscopic images to examine the efficacy of different quantitative approaches for characterizing phytoliths. A comprehensive experiment performed on images of 429 phytoliths demonstrated that the automatic phytolith classification is a promising area of research that will help researchers to invest time more efficiently and improve their recognition accuracy rate.
Plant phenotyping relevance
植物由来の植物珪酸体を対象に、顕微鏡画像からの自動識別・分類手法を比較評価しており、画像ベースの形態的特徴抽出が研究の中心です。
abstractThis paper presents a comparative analysis of six classification methods, using digitized microscopic images to examine the efficacy of different quantitative approaches for characterizing phytoliths.
abstractA comprehensive experiment performed on images of 429 phytoliths demonstrated that the automatic phytolith classification is a promising area of research
Code and data availability
The paper's phytolith photomicrograph dataset (429 images across 8 morphotypes) is explicitly stated to be publicly available at the UPF repository, and the authors' analysis code is shared on GitHub with explicit availability language. Both are paper-specific, public, and actionable.
classification process. Even though several researchers have attempted to create automatic tools for the identification of archaeobotanical remains, none of the attempts has produced a tool that is accessible online or as a downloadable app. We are sharing the code used in our research (which is accessible at https://github.com/alvarag/AutomaticPhytolithClassification) to stimulate other researchers to join in the effort to build a real and functional tool that can be trained online, increasing its accuracy. Future Research Lines The development of new features and the application of feature selection techniques are some of the research avenues we are plan- ning to explore. It would be impor
Open resource ↗github.com/alvarag/AutomaticPhytolithClassification · pdf-raw-page:9 lines:1-85This is an automatically classified, unverified record. Curator approval is required before any resource enters the Catalog.