Image Analysis in Plant Sciences: Publish Then Perish. (July 2017)

• Record Type: Journal Article
• Authors: Lobet, Guillaume
• Is Part Of: Trends in plant science. Volume 22:Number 7(2017)
• Page Start: 559
• Page End: 566
• Publication Date: 2017-07
• Subjects: image analysis -- open science -- open data -- validation -- maintenance; Botany -- Periodicals; Botanique -- Périodiques; Botany; Periodicals; 580.5
• DOI: 10.1016/j.tplants.2017.05.002
• Languages: English
• ISSNs: 1360-1385

Abstract

Abstract : Image analysis has become a powerful technique for most plant scientists. In recent years dozens of image analysis tools have been published in plant science journals. These tools cover the full spectrum of plant scales, from single cells to organs and canopies. However, the field of plant image analysis remains in its infancy. It still has to overcome important challenges, such as the lack of robust validation practices or the absence of long-term support. In this Opinion article, I: (i) present the current state of the field, based on data from the plant-image-analysis.org database; (ii) identify the challenges faced by its community; and (iii) propose workable ways of improvement. Trends: Image analysis techniques and tools are used by a large proportion of the plant science community. Existing tools address a large proportion of current needs. Validation and benchmarking of these tools is variable and often overlooked. Long-term maintenance of these tools is not always ensured.