Transfer Learning from Synthetic Data Applied to Soil–Root Segmentation in X-Ray Tomography Images
| Autor: | Richard Schielein, Stefan Gerth, David Rousseau, Carole Frindel, Clément Douarre |
|---|---|
| Přispěvatelé: | Université d'Angers (UA), Fraunhofer Institute for Systems and Innovation Research (Fraunhofer ISI ), Fraunhofer (Fraunhofer-Gesellschaft), Université Claude Bernard Lyon 1 (UCBL), Université de Lyon, Institut de Recherche en Horticulture et Semences (IRHS), Université d'Angers (UA)-Institut National de la Recherche Agronomique (INRA)-AGROCAMPUS OUEST, Institut national d'enseignement supérieur pour l'agriculture, l'alimentation et l'environnement (Institut Agro)-Institut national d'enseignement supérieur pour l'agriculture, l'alimentation et l'environnement (Institut Agro), French Government : ANR-11-BTBR-0007, EU COST action 'The quest for tolerant varieties: phenotyping at plant and cellular level' FA1306. |
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
root systems
segmentation X-ray tomography transfer learning 0106 biological sciences [SDV.SA]Life Sciences [q-bio]/Agricultural sciences Computer science ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION 02 engineering and technology 01 natural sciences Synthetic data GeneralLiterature_MISCELLANEOUS Low contrast Robustness (computer science) 0202 electrical engineering electronic engineering information engineering Traitement du signal et de l'image [SDV.BV]Life Sciences [q-bio]/Vegetal Biology Radiology Nuclear Medicine and imaging Segmentation Electrical and Electronic Engineering agriculture Vegetal Biology business.industry Deep learning Signal and Image processing Pattern recognition relation plante-sol Computer Graphics and Computer-Aided Design Agricultural sciences Simulated data 020201 artificial intelligence & image processing Computer Vision and Pattern Recognition Artificial intelligence Tomography Transfer of learning business [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing Biologie végétale Sciences agricoles 010606 plant biology & botany |
| Zdroj: | Journal of Imaging Journal of Imaging, MDPI, 2018, 4 (5), pp.65. ⟨10.3390/jimaging4050065⟩ www.mdpi.com/journal/jimaging Journal of Imaging; Volume 4; Issue 5; Pages: 65 Journal of Imaging 5 (4), 65. (2018) |
| ISSN: | 2313-433X |
| DOI: | 10.3390/jimaging4050065⟩ |
| Popis: | International audience; One of the most challenging computer vision problems in the plant sciences is the segmentation of roots and soil in X-ray tomography. So far, this has been addressed using classical image analysis methods. In this paper, we address this soil-root segmentation problem in X-ray tomography using a variant of supervised deep learning-based classification called transfer learning where the learning stage is based on simulated data. The robustness of this technique, tested for the first time with this plant science problem, is established using soil-roots with very low contrast in X-ray tomography. We also demonstrate the possibility of efficiently segmenting the root from the soil while learning using purely synthetic soil and roots. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|