processing, reviewing, curating, describing, and hosting the data. Instead, we focused on an initial public release and plan to make new datasets available based on need. Access to unpublished data can be requested from the authors, and as data are curated they will be added to subsequent versions of the public domain release ( https://terraref.org/data/access-data ). In addition to hosting an archival copy of data on Dryad [ 16 ] , the documentation includes instructions for browsing and accessing these data through a variety of online portals. These portals provide access to web user interfaces as well as databases, APIs, and R and Python clients. In some cases it will be easier to acce
Open resource ↗lines:234-317Unverified paper record
What Does TERRA-REF's High Resolution, Multi Sensor Plant Sensing Public Domain Data Offer the Computer Vision Community?
arXiv · 29 Jul 2021 · 10.48550/arxiv.2107.14072
Abstract
A core objective of the TERRA-REF project was to generate an open-access reference dataset for the evaluation of sensing technologies to study plants under field conditions. The TERRA-REF program deployed a suite of high-resolution, cutting edge technology sensors on a gantry system with the aim of scanning 1 hectare (10$^4$) at around 1 mm$^2$ spatial resolution multiple times per week. The system contains co-located sensors including a stereo-pair RGB camera, a thermal imager, a laser scanner to capture 3D structure, and two hyperspectral cameras covering wavelengths of 300-2500nm. This sensor data is provided alongside over sixty types of traditional plant phenotype measurements that can be used to train new machine learning models. Associated weather and environmental measurements, information about agronomic management and experimental design, and the genomic sequences of hundreds of plant varieties have been collected and are available alongside the sensor and plant phenotype data. Over the course of four years and ten growing seasons, the TERRA-REF system generated over 1 PB of sensor data and almost 45 million files. The subset that has been released to the public domain accounts for two seasons and about half of the total data volume. This provides an unprecedented opportunity for investigations far beyond the core biological scope of the project. The focus of this paper is to provide the Computer Vision and Machine Learning communities an overview of the available data and some potential applications of this one of a kind data.
Plant phenotyping relevance
植物の高解像度マルチセンサーデータと植物表現型データを含む公開ベンチマーク／データセットを紹介し、コンピュータビジョンでの利用を主目的とするため、フェノタイピング手法・基盤として中心的です。
abstractgenerate an open-access reference dataset for the evaluation of sensing technologies to study plants under field conditions
abstractThe system contains co-located sensors including a stereo-pair RGB camera, a thermal imager, a laser scanner to capture 3D structure, and two hyperspectral cameras covering wavelengths of 300-2500nm.
abstractThis sensor data is provided alongside over sixty types of traditional plant phenotype measurements that can be used to train new machine learning models.
Code and data availability
The paper describes the TERRA-REF public domain release of plant phenotyping sensor data (RGB, thermal, laser scanner, hyperspectral, PSII) plus derived phenotypes, and explicitly points to public code repositories for the processing pipeline (terraref GitHub, PhytoOracle, AgPipeline) and a data access portal. All are,
approach described by Li et al . [ 18 ] . Herritt et al . [ 14 , 13 ] demonstrate and provide software used in analysis of a sequence of images that capture plant fluorescence response to a pulse of light. Most of the algorithms used to generate data products have not been published as papers but are made available on GitHub ( https://github.com/terraref ); code used to release the data publication in 2020 is available on Zenodo [ 25 , 15 , 10 , 6 , 4 , 19 , 8 , 7 , 5 , 9 , 17 ] . Pipeline development continues to support ongoing use of the field scanner as well as more general applications in plant sensing pipelines. Recent advances have improved pipeline scalability and modul
Open resource ↗terraref · lines:193-233lant sensing pipelines. Recent advances have improved pipeline scalability and modularity by adopting workflow tools and making use of heterogeneous computing environments. The TERRA-REF computing pipeline has been adapted and extended for continuing use with the Field Scanner with the new name ”PhytoOracle” and is available at https://github.com/LyonsLab/PhytoOracle . Related work generalizing the pipeline for other phenomics applications has been released under the name ”AgPipeline” https://github.com/agpipeline with applications to aerial imaging described by Schnaufer et al . [ 22 ] . All of these software are made available with permissive open source licenses on GitHub to enable acces
Open resource ↗PhytoOracle · lines:193-233nvironments. The TERRA-REF computing pipeline has been adapted and extended for continuing use with the Field Scanner with the new name ”PhytoOracle” and is available at https://github.com/LyonsLab/PhytoOracle . Related work generalizing the pipeline for other phenomics applications has been released under the name ”AgPipeline” https://github.com/agpipeline with applications to aerial imaging described by Schnaufer et al . [ 22 ] . All of these software are made available with permissive open source licenses on GitHub to enable access and community development. Figure 4: Summary of public sensor datasets from Seasons 4 and 6. Each dot represents the dates for which a particular da
Open resource ↗agpipeline · lines:193-233This is an automatically classified, unverified record. Curator approval is required before any resource enters the Catalog.