Method development by precision phenotyping coupled with automated data acquisition and integrated data management
A transformative discovery platform for ground based field phenotyping, TraitSeeker is currently being adopted by the world’s leading agricultural research institutions.
The TraitSeeker payload delivers a large and comprehensive set of high-resolution sensors that provides high quality/ high dimensionality data regardless of ambient conditions which is the optimal starting point for developing new methods for plant sciences and breeding.
The core of TraitSeeker is an industrial data acquisition sensor fusion framework designed to automate a-synchronous acquisition of imaging data. It combines with ultra-high precision timing technology to co-register acquired data with metadata for pixel level fusion- unique in this application and resulting in ultra-rich 3D datasets.
TraitSeeker is a fully integrated turn-key ground based field phenotyping platform complementary with aerial platforms. Of rugged agricultural grade construction and industrial components, TraitSeeker is an AWD platform designed for durability in adverse testing conditions and rapid, easy deployment across multiple trial sites by the operator. Designed for technological enhancements over ensuing years, the TraitSeeker approach delivers efficiency and affordability for multi-trial field phenotyping research and breeding programs.
Unique in its ability to supply its own full bandwidth illumination supporting VNIR and SWIR hyperspectral imaging in changing ambient conditions, TraitSeeker also provides LiDAR, RGB, INS and environmental sensors to provide comprehensive detailed information on physiological processes, biochemical traits responses and plant breeding beyond the capabilities of aerial platforms.
Autonomous drive is coupled with automated plot segmentation. Only plot data is stored from selected sensors and autonomously stored for automated output of a variety of plot analytics using validated algorithms. Onboard sensor calibration is provided and a tiered suite of tools provide status updates of the system performance.
Accelerating development
discover traits critical for
research or breeding
If image-based phenotyping is critical TraitSeeker offers automated enhanced data acquisition and processing
Row Labels | Average Total Area (m2) | Average Crop Height (m) | NDVI | Average Ground Cover (%) | Averge Total Dig. Biomass (m3) |
---|---|---|---|---|---|
Iceberg | 100% | 100% | 100% | 100% | 100% |
1 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
2 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
3 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
4 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
5 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
6 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
7 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
8 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
9 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
10 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
11 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
12 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
13 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
14 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
15 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
16 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
17 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
18 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
19 | 99.98% | 97.72% | 100.77% | 103.83% | 101.62% |
20 | 100.98% | 99.48% | 100.66% | 103.99% | 104.42% |
Excellent repeatability and accuracy
Data handling is often considered a major obstacle in high throughput field phenotyping. Furthermore, traditional measurements are subjective, suffer from human error and repeatability affecting trait heritability. Overcoming these hurdles with standardized datasets is a critical feature to reducing labor intensity and cost reduction in research and breeding programs.
TraitSeeker provides raw plot-only data or has its own pipeline to automatically transform data. Data processing is automated, user-friendly, intuitive and rapid; converting sensor information into annotated, georeferenced plot specific data including quantitative trait metrics and plot trait images. Centralized storage means multi-plot/ multi-site analyses can be performed concomitantly.
Individual plants can also be identified at emergence, tracked and temporally assessed.
TraitSeeker produces metadata enriched datasets that are Deep Learning and Crop Growth Model compatible, according to FAIR principles, MIAPPE compliant and suited to ontological approaches such as PHIS.
All data is saved directly into annotated plot files and includes csv files of plot specific data + individual plant data, fused/non-fused data, image parameters, high-res stitched plot RGB, high-res 3D files, metadata, calibration and event logs. Data can be transferred to centralized processing via SSDs networked to the onboard industrial PC.
Making acquisition of large
phenotypic data sets affordable
For plant breeders interested in image based phenotyping CropTraits is developing a specific solution for high throughput phenotyping based on affordable, simultaneous and quantitative measurements at different locations.
TraitSeeker – chosen by world-leading research institutes
PhenoKey
TraitSeeker is the result of an ongoing collaboration between CropTraits and PhenoKey over several years.