Towards Low-Cost Hyperspectral Single-Pixel Imaging for Plant Phenotyping

Ribes, Mathieu; Russias, Gaspard; Trégoat, Denis; Fournier, Antoine

doi:10.3390/s20041132

Cited by 7 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, hyperspectral imaging has gained importance in recent years. However, since cost-effective applications are not yet available, research continues on systems with only one measurement point for plant characterization [218]. All of those systems generate large amounts of data.…”

Section: Farmer-in-the-loopmentioning

confidence: 99%

Digital Transformation in Smart Farm and Forest Operations Needs Human-Centered AI: Challenges and Future Directions

Holzinger

Saranti

Angerschmid

et al. 2022

Sensors

View full text Add to dashboard Cite

The main impetus for the global efforts toward the current digital transformation in almost all areas of our daily lives is due to the great successes of artificial intelligence (AI), and in particular, the workhorse of AI, statistical machine learning (ML). The intelligent analysis, modeling, and management of agricultural and forest ecosystems, and of the use and protection of soils, already play important roles in securing our planet for future generations and will become irreplaceable in the future. Technical solutions must encompass the entire agricultural and forestry value chain. The process of digital transformation is supported by cyber-physical systems enabled by advances in ML, the availability of big data and increasing computing power. For certain tasks, algorithms today achieve performances that exceed human levels. The challenge is to use multimodal information fusion, i.e., to integrate data from different sources (sensor data, images, *omics), and explain to an expert why a certain result was achieved. However, ML models often react to even small changes, and disturbances can have dramatic effects on their results. Therefore, the use of AI in areas that matter to human life (agriculture, forestry, climate, health, etc.) has led to an increased need for trustworthy AI with two main components: explainability and robustness. One step toward making AI more robust is to leverage expert knowledge. For example, a farmer/forester in the loop can often bring in experience and conceptual understanding to the AI pipeline—no AI can do this. Consequently, human-centered AI (HCAI) is a combination of “artificial intelligence” and “natural intelligence” to empower, amplify, and augment human performance, rather than replace people. To achieve practical success of HCAI in agriculture and forestry, this article identifies three important frontier research areas: (1) intelligent information fusion; (2) robotics and embodied intelligence; and (3) augmentation, explanation, and verification for trusted decision support. This goal will also require an agile, human-centered design approach for three generations (G). G1: Enabling easily realizable applications through immediate deployment of existing technology. G2: Medium-term modification of existing technology. G3: Advanced adaptation and evolution beyond state-of-the-art.

show abstract

Section: Farmer-in-the-loopmentioning

confidence: 99%

Digital Transformation in Smart Farm and Forest Operations Needs Human-Centered AI: Challenges and Future Directions

Holzinger

Saranti

Angerschmid

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Moreover, even if they are equipped with simple RGB sensors, professional automated phenotyping platforms can be too expensive for smaller companies. As a solution, several studies have presented low-cost systems for plant phenotyping ( Bagley et al , 2020 ; Ribes et al , 2020 ; Wu et al , 2020 ). However, such systems can serve well for scientific purposes but do not necessarily need to meet the routine and standardized screening campaigns called for in industrial practices.…”

Section: Indoor Phenotyping For Plant Biostimulant Screening and Stud...mentioning

confidence: 99%

Presence and future of plant phenotyping approaches in biostimulant research and development

Diego

Spíchal

2022

Journal of Experimental Botany

View full text Add to dashboard Cite

Commercial interest in biostimulants as a tool of sustainable green economics and agriculture concepts is on the steep increase, being followed by increasing demand to employ efficient scientific ways to develop new products and understand their mechanisms of action. Biostimulants represent a highly diverse group of agents derived from various natural sources. Regardless of their nutrition content and composition, they are classified by their ability to improve crop performance through enhanced nutrient use efficiency, abiotic stress tolerance, and quality of crops. Numerous reports have described modern, non-invasive sensor-based phenotyping methods in plant research. This review focuses on applying phenotyping approaches in biostimulant research and development and maps the evolution of interaction of these two intensively growing domains. How phenotyping served to identify new biostimulants, the description of their biological activity, and the mechanism/mode of action are summarized. Special attention is dedicated to the indoor high-throughput methods using model plants suitable for biostimulant screening and developmental pipelines and high-precision approaches used to determine biostimulant activity. The need for a complex way of testing biostimulants as multi-component products through integrating other –omic approaches followed by advanced statistical/mathematical tools is emphasized.

show abstract

“…Recently, hyperspectral imaging systems working in the visible 11,12 or in the near-infrared regions 13 have been reported. They are based on single-pixel imaging 14 where a sequence of illumination patterns is superimposed on the scene and the averaged intensity is recorded by a detector.…”

Section: Introductionmentioning

confidence: 99%

Spatial-frequency-domain hyperspectral microscopy

Uguen,

Piedevache,

Russias

et al. 2024

Unconventional Optical Imaging IV

Self Cite

View full text Add to dashboard Cite

Hyperspectral imaging allows to reconstruct both spatial and quasi-continuous spectral information of an object (i.e., an hypercube), unlike multispectral imaging where the spectrum is divided in few bands. This work shows the innovating combination of single-pixel hyperspectral imaging with reflective-mode optical microscopy to retrieve high-spatial-resolution hypercubes of samples in the visible and in the near-infrared ranges. The spatial resolution reaches around 2.5 µm and the spectral information is sampled over around 200 spectral bands. Furthermore, statistical algorithms enables the clustering of regions of interest of inorganic and organic samples and the identification of spectral responses of targeted cells.

show abstract

Towards Low-Cost Hyperspectral Single-Pixel Imaging for Plant Phenotyping

Cited by 7 publications

References 19 publications

Digital Transformation in Smart Farm and Forest Operations Needs Human-Centered AI: Challenges and Future Directions

Digital Transformation in Smart Farm and Forest Operations Needs Human-Centered AI: Challenges and Future Directions

Presence and future of plant phenotyping approaches in biostimulant research and development

Spatial-frequency-domain hyperspectral microscopy

Contact Info

Product

Resources

About