Machine Learning for Plant Phenotyping Needs Image Processing

Abstract

We found the article by Singh et al. [1] extremely interesting since it introduces and showcases the utility of machine learning for high throughput data-driven plant phenotyping.With this letter we want to emphasize the role that image analysis and processing have in the phenotyping pipeline beyond what [1] suggests, both in analyzing phenotyping data (e.g., to measure growth) but also 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] consider as part of pre-processing) that has brought a renaissance in phenotyping [2]. At the same time, the lack of robust methods to analyze these images is now the new bottleneck [3][4][5]. And this bottleneck is not easy to overcome. As the following aims to illustrate, it is coupled to the imaging system and the environment but also to the analysis task at hand and requires new skills to help deal with the challenges introduced.A successful high-throughput image-based phenotyping system starts with the imaging approach itself. The choices are to image many plants simultaneously or one plant at a time,

Deep Plant Phenomics: A Deep Learning Platform for Complex Plant Phenotyping Tasks

“…It has been proposed that future progress in image-based plant phenotyping will require a combined effort in the domains of image processing for feature extraction and machine learning for data analysis (Tsaftaris et al, 2016). In the current machine learning literature, deep learning methods lead the state of the art in many image-based tasks such as object detection and localization, semantic segmentation, image classification, and others (LeCun et al, 2015).…”

“…In response to the limited flexibility and poor performance of classical image processing pipelines for complex phenotyping tasks, machine learning techniques are expected to take a prominent role in the future of image-based phenotyping (Tsaftaris et al, 2016). Plant disease detection and diagnosis is an example of a complex phenotyping task where machine learning techniques, such as support vector machines, clustering algorithms, and neural networks, have demonstrated success (Singh et al, 2016).…”

“…Given the success in other areas, deep learning has been proposed as a future trend in image-based plant phenotyping (Tsaftaris et al, 2016). Early results from the few studies that have applied the technique are promising: CNNs were effective for plant disease detection and diagnosis (Mohanty et al, 2016) and for classifying fruits and flowers of plants in field images (Pawara et al, 2017).…”

Deep Learning for Multi-task Plant Phenotyping

“…Of course, machine learning, and especially deep machine learning approaches are fuelled only by high quality, annotated datasets [24,15]. For learning to be effective and efficient, the image data the computer is learning from must be both accurately captured and well annotated.…”

Evaluation of 3D/2D Imaging and Image Processing Techniques for the Monitoring of Seed Imbibition

“…The new bottleneck for these questions is now moving in the direction of image processing in order to efficiently and automatically extract quantitative phenotypic traits from the acquired data [8]. In this perspective, emerging works exploit advanced machine learning techniques to automatically determine the best feature spaces enabling one to address given informational tasks [9]. However, these image processing approaches are especially efficient, in terms of computation time or learning dataset size, when applied to well-contrasted imaging.…”