Root identification in minirhizotron imagery with multiple instance learning

Yu, Guohao; Zare, Alina; Sheng, Hudanyun; Matamala, Roser; Reyes-Cabrera, Joel; Fritschi, Felix B.; Juenger, Thomas E.

doi:10.1007/s00138-020-01088-z

Root identification in minirhizotron imagery with multiple instance learning

Journal Article · Wed Jun 24 00:00:00 EDT 2020 · Machine Vision and Applications

DOI:https://doi.org/10.1007/s00138-020-01088-z· OSTI ID:1773853

^[1]; Zare, Alina ^[2]; Sheng, Hudanyun ^[2]; Matamala, Roser ^[3]; Reyes-Cabrera, Joel ^[4]; Fritschi, Felix B. ^[4]; Juenger, Thomas E. ^[5]

Univ. of Florida, Gainesville, FL (United States); University of Texas at Austin
Univ. of Florida, Gainesville, FL (United States)
Argonne National Lab. (ANL), Lemont, IL (United States)
Univ. of Missouri, Columbia, MO (United States)
Univ. of Texas, Austin, TX (United States)

In this study, multiple instance learning (MIL) algorithms to automatically perform root detection and segmentation in minirhizotron imagery using only image-level labels are proposed. Root and soil characteristics vary from location to location, and thus, supervised machine learning approaches that are trained with local data provide the best ability to identify and segment roots in minirhizotron imagery. However, labeling roots for training data (or otherwise) is an extremely tedious and time-consuming task. This paper aims to address this problem by labeling data at the image level (rather than the individual root or root pixel level) and train algorithms to perform individual root pixel level segmentation using MIL strategies. Three MIL methods (multiple instance adaptive cosine coherence estimator, multiple instance support vector machine, multiple instance learning with randomized trees) were applied to root detection and compared to non-MIL approaches. The results show that MIL methods improve root segmentation in challenging minirhizotron imagery and reduce the labeling burden. In our results, multiple instance support vector machine outperformed other methods. The multiple instance adaptive cosine coherence estimator algorithm was a close second with an added advantage that it learned an interpretable root signature which identified the traits used to distinguish roots from soil and did not require parameter selection.

View Accepted Manuscript (DOE)

Research Organization:: Univ. of Texas, Austin, TX (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Biological and Environmental Research (BER); USDOE Advanced Research Projects Agency - Energy (ARPA-E)

Grant/Contract Number:: SC0014156; AR0000820

OSTI ID:: 1773853

Journal Information:: Machine Vision and Applications, Journal Name: Machine Vision and Applications Journal Issue: 6 Vol. 31; ISSN 0932-8092

Publisher:: SpringerCopyright Statement

Country of Publication:: United States

Language:: English

References (33)

Fine root biomass and turnover of two fast-growing poplar genotypes in a short-rotation coppice culture Berhongaray, Gonzalo; Janssens, I. A.; King, J. S. Plant and Soil, Vol. 373, Issue 1-2 https://doi.org/10.1007/s11104-013-1778-x	journal	June 2013
Solving the multiple instance problem with axis-parallel rectangles Dietterich, Thomas G.; Lathrop, Richard H.; Lozano-Pérez, Tomás Artificial Intelligence, Vol. 89, Issue 1-2 https://doi.org/10.1016/s0004-3702(96)00034-3	journal	January 1997
Net energy of cellulosic ethanol from switchgrass Schmer, M. R.; Vogel, K. P.; Mitchell, R. B. Proceedings of the National Academy of Sciences, Vol. 105, Issue 2 https://doi.org/10.1073/pnas.0704767105	journal	January 2008
Pattern Recognition Using Functions of Multiple Instances Zare, Alina; Gader, Paul 2010 20th International Conference on Pattern Recognition (ICPR) https://doi.org/10.1109/icpr.2010.273	conference	August 2010
Hybrid Detectors for Subpixel Targets Broadwater, Joshua; Chellappa, Rama IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 29, Issue 11 https://doi.org/10.1109/tpami.2007.1104	journal	November 2007
Observation of plant roots in situ Waddington, J. Canadian Journal of Botany, Vol. 49, Issue 10 https://doi.org/10.1139/b71-261	journal	October 1971
A New Method for Root Detection in Minirhizotron Images: Hypothesis Testing Based on Entropy-based Geometric Level Set Decision Shojaedini, S. V. International Journal of Engineering, Vol. 27, Issue 1 (A) https://doi.org/10.5829/idosi.ije.2014.27.01a.12	journal	November 2013
Stacked Hourglass Networks for Human Pose Estimation Newell, Alejandro; Yang, Kaiyu; Deng, Jia Computer Vision – ECCV 2016 https://doi.org/10.1007/978-3-319-46484-8_29	book	January 2016
Detecting and Measuring Fine Roots in Minirhizotron Images Using Matched Filtering and Local Entropy Thresholding Zeng, Guang; Birchfield, Stanley T.; Wells, Christina E. Machine Vision and Applications, Vol. 17, Issue 4 https://doi.org/10.1007/s00138-006-0024-4	journal	July 2006
Rapid automated detection of roots in minirhizotron images Zeng, Guang; Birchfield, Stanley T.; Wells, Christina E. Machine Vision and Applications, Vol. 21, Issue 3 https://doi.org/10.1007/s00138-008-0179-2	journal	December 2008
Generalized Dictionaries for Multiple Instance Learning Shrivastava, Ashish; Patel, Vishal M.; Pillai, Jaishanker K. International Journal of Computer Vision, Vol. 114, Issue 2-3 https://doi.org/10.1007/s11263-015-0831-z	journal	June 2015
Solving the multiple instance problem with axis-parallel rectangles Dietterich, Thomas G.; Lathrop, Richard H.; Lozano-Pérez, Tomás Artificial Intelligence, Vol. 89, Issue 1-2 https://doi.org/10.1016/S0004-3702(96)00034-3	journal	January 1997
Advancing fine root research with minirhizotrons Johnson, M. G.; Tingey, D. T.; Phillips, D. L. Environmental and Experimental Botany, Vol. 45, Issue 3 https://doi.org/10.1016/S0098-8472(01)00077-6	journal	June 2001
Dairy goat detection based on Faster R-CNN from surveillance video Wang, Dong; Tang, JingLei; Zhu, Weijie Computers and Electronics in Agriculture, Vol. 154 https://doi.org/10.1016/j.compag.2018.09.030	journal	November 2018
Overcoming small minirhizotron datasets using transfer learning Xu, Weihuang; Yu, Guohao; Zare, Alina Computers and Electronics in Agriculture, Vol. 175 https://doi.org/10.1016/j.compag.2020.105466	journal	August 2020
Random set framework for multiple instance learning Bolton, Jeremy; Gader, Paul; Frigui, Hichem Information Sciences, Vol. 181, Issue 11 https://doi.org/10.1016/j.ins.2010.12.020	journal	June 2011
Variation in root architecture among switchgrass cultivars impacts root decomposition rates Graaff, Marie-Anne de; Six, Johan; Jastrow, Julie D. Soil Biology and Biochemistry, Vol. 58 https://doi.org/10.1016/j.soilbio.2012.11.015	journal	March 2013
Random Forests Breiman, Leo Machine Learning, Vol. 45, Issue 1, p. 5-32 https://doi.org/10.1023/A:1010933404324	journal	January 2001
A Device for the Observation of Root Growth in the Soil Bates, G. H. Nature, Vol. 139, Issue 3527 https://doi.org/10.1038/139966b0	journal	June 1937
Root Morphological Traits and Spatial Distribution under Different Nitrogen Treatments and Their Relationship with Grain Yield in Super Hybrid Rice Liu, Ke; He, Aibin; Ye, Chang Scientific Reports, Vol. 8, Issue 1 https://doi.org/10.1038/s41598-017-18576-4	journal	January 2018
RootNav 2.0: Deep learning for automatic navigation of complex plant root architectures Yasrab, Robail; Atkinson, Jonathan A.; Wells, Darren M. GigaScience, Vol. 8, Issue 11 https://doi.org/10.1093/gigascience/giz123	journal	November 2019
The CFAR adaptive subspace detector is a scale-invariant GLRT Kraut, S.; Scharf, L. L. IEEE Transactions on Signal Processing, Vol. 47, Issue 9 https://doi.org/10.1109/78.782198	journal	January 1999
Adaptive subspace detectors Kraut, S.; Scharf, L. L.; McWhorter, L. T. IEEE Transactions on Signal Processing, Vol. 49, Issue 1 https://doi.org/10.1109/78.890324	journal	January 2001
Dictionary-based multiple instance learning Shrivastava, Ashish; Pillai, Jaishanker K.; Patel, Vishal M. 2014 IEEE International Conference on Image Processing (ICIP) https://doi.org/10.1109/ICIP.2014.7025031	conference	October 2014
Pattern Recognition Using Functions of Multiple Instances Zare, Alina; Gader, Paul 2010 20th International Conference on Pattern Recognition (ICPR) https://doi.org/10.1109/ICPR.2010.273	conference	August 2010
Functions of Multiple Instances for Learning Target Signatures Jiao, Changzhe; Zare, Alina IEEE Transactions on Geoscience and Remote Sensing, Vol. 53, Issue 8 https://doi.org/10.1109/TGRS.2015.2406334	journal	August 2015
Remote Sensing Image Stripe Noise Removal: From Image Decomposition Perspective Chang, Yi; Yan, Luxin; Wu, Tao IEEE Transactions on Geoscience and Remote Sensing, Vol. 54, Issue 12 https://doi.org/10.1109/TGRS.2016.2594080	journal	December 2016
Hybrid Detectors for Subpixel Targets Broadwater, Joshua; Chellappa, Rama IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 29, Issue 11 https://doi.org/10.1109/TPAMI.2007.1104	journal	November 2007
SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation Badrinarayanan, Vijay; Kendall, Alex; Cipolla, Roberto IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 39, Issue 12 https://doi.org/10.1109/TPAMI.2016.2644615	journal	December 2017
Discriminative Multiple Instance Hyperspectral Target Characterization Zare, Alina; Jiao, Changzhe; Glenn, Taylor IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 40, Issue 10 https://doi.org/10.1109/TPAMI.2017.2756632	journal	October 2018
Segmentation of roots in soil with U-Net Smith, Abraham George; Petersen, Jens; Selvan, Raghavendra Plant Methods, Vol. 16, Issue 1 https://doi.org/10.1186/s13007-020-0563-0	journal	February 2020
Can Changes in Canopy and/or Root System Architecture Explain Historical Maize Yield Trends in the U.S. Corn Belt? Hammer, Graeme L.; Dong, Zhanshan; McLean, Greg Crop Science, Vol. 49, Issue 1 https://doi.org/10.2135/cropsci2008.03.0152	journal	January 2009
Root traits contributing to plant productivity under drought Comas, Louise H.; Becker, Steven R.; Cruz, Von Mark V. Frontiers in Plant Science, Vol. 4 https://doi.org/10.3389/fpls.2013.00442	journal	January 2013

Similar Records

Weakly Supervised Minirhizotron Image Segmentation with MIL-CAM

Journal Article · Mon Jan 04 23:00:00 EST 2021 · Lecture Notes in Computer Science · OSTI ID:2448006

Overcoming small minirhizotron datasets using transfer learning

Journal Article · Fri Jun 19 00:00:00 EDT 2020 · Computers and Electronics in Agriculture · OSTI ID:1773850

PRMI: A Dataset of Minirhizotron Images for Diverse Plant Root Study

Conference · Fri Dec 31 23:00:00 EST 2021 · OSTI ID:1887572

Related Subjects

59 BASIC BIOLOGICAL SCIENCES
image processing
image segmentation
imprecise labels
machine learning
minirhizotron
minirhizotron image
multiple instance learning
plant roots
roots

Root identification in minirhizotron imagery with multiple instance learning

Citation Formats

References (33)

Similar Records

Related Subjects