Machine Learning for Plant Phenotyping Needs Image Processing

We found the article by Singh et al. [ 1 Singh A. et al. Machine learning for high-throughput stress phenotyping in plants. Trends Plant Sci. 2016; 21: 110-124 Abstract Full Text Full Text PDF PubMed Scopus (506) Google Scholar ] extremely interesting because it introduces and showcases the utility of machine learning for high-throughput data-driven plant phenotyping. With this letter we aim to emphasize the role that image analysis and processing have in the phenotyping pipeline beyond what is suggested in [ 1 Singh A. et al. Machine learning for high-throughput stress phenotyping in plants. Trends Plant Sci. 2016; 21: 110-124 Abstract Full Text Full Text PDF PubMed Scopus (506) Google Scholar ], both in analyzing phenotyping data (e.g., to measure growth) and when providing effective feature extraction to be used by machine learning. Key recent reviews have shown that it is image analysis itself (what the authors of [ 1 Singh A. et al. Machine learning for high-throughput stress phenotyping in plants. Trends Plant Sci. 2016; 21: 110-124 Abstract Full Text Full Text PDF PubMed Scopus (506) Google Scholar ] consider as part of pre-processing) that has brought a renaissance in phenotyping [ 2 Spalding E.P. Miller N.D. Image analysis is driving a renaissance in growth measurement. Curr. Opin. Plant Biol. 2013; 16: 100-104 Crossref PubMed Scopus (79) Google Scholar ]. At the same time, the lack of robust methods to analyze these images is now the new bottleneck [ 3 Minervini M. et al. Image analysis: the new bottleneck in plant phenotyping (applications corner). IEEE Signal Process. Mag. 2015; 32: 126-131 Crossref Scopus (157) Google Scholar , 4 Rousseau D. et al. Imaging methods for phenotyping of plant traits. in: Kumar J. Phenomics in Crop Plants: Trends, Options and Limitations. Springer India, 2015: 61-74 Crossref Scopus (18) Google Scholar , 5 Pridmore T.P. et al. What lies beneath: underlying assumptions in bioimage analysis. Trends Plant Sci. 2012; 17: 688-692 Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar ] – and this bottleneck is not easy to overcome. As the following aims to illustrate, it is coupled not only to the imaging system and the environment but also to the analytical task at hand, and requires new skills to help deal with the challenges introduced.