The fastest digital seed testing system measuring seed, germination, and seedling emergence quality.
The LemnaTec SeedAIxpert HT is a phenotyping system for seeds and seedlings. It consists of an imaging unit with a cameras and lights that moves along shelf systems. Seed trays located in the shelves are automatically moved in front of the camera lens and imaged. The imaging unit is connected to a computer system where images are analyzed with advanced image processing algorithms including machine learning. Key application is the assessment of seeds germinating on moist paper, which includes phenotypic assessments of seedlings. Further applications include seedling phenotyping or seedling quality assessments.
click image to enlarge
Features
Application Examples
Germination detection
Roots emerging from seeds were detected by SeedAIxpert and marked in yellow. Germination percentage (seeds with / without roots in relation to total seed number) is provided. Detected roots can be measured, e.g. root lenght.
Germination detection of coated seeds
SeedAIxpert is designed to work with coated seeds, too. Here coated seeds were classified as germinated or not germinated according to the occurence of roots. As SeedAIxpert can work with color analysis, it can handle differently coated seeds at once.
SeedAIxpert works with most seeds
SeedAIxpert technology is able to cope with seeds of different sizes and colors. No matter whether you work with small herb or vegetable seeds or with maize, SeedAIxpert makes seed testing easy.
Artificial intelligence for seedling classification
Emerging seedlings consist of various parts, e.g. roots, shoots, or root hairs. With machine learning algorithms, SeedAIxpert can be trained to recognize each of these parts separately.
Maize germination
Example for a maize germination test as top-of-paper assay. SeedAIxpert images the samples and derives germination quality data from the images.
Vegetable seed germination classified with machine learning
Vegetables germinating in a top-of-paper test were classified with the SeedAIxpert machine learning algorithms. Occurence and size of shoot and root are analyzed and a subsequent quality rating, e.g. as usable plant test can be set up.
Seed classification
Embryo and Endosperm are visible on the surface of maize seeds. Imaged and classified by SeedAIxpert, fractions of embryo and endosperm are measured for each seed.
Seedling emergence test
Oilseed rape germination was assessed in a seedling emergence test with seeds placed in substrate. Emerging cotyledons were recognized and measured for germination frequency and seedling quality.
Seed testing can be laborious and frequently relies on old-fashioned scoring procedures. With our new LemnaTec SeedAIxpert we are bringing the fastest digital seed germination and seedling emergence test into the growing market of high-quality seeds. It saves time and labor by automatically analyzing seed germination percentage and timing together with metrics on the seeds and seedlings. The digital metrics give an advantage compared to counting and visual scoring the seed germination assays.
Key advantages are:
- Rapid digital analysis of seed germination or seedling emergence
- Saving time and labour
- Improved data quality
- Broad range of seed types
- Seed and seedling quality traits
- Normal/abnormal seedling and usable plant determination (optional advanced analysis feature)
- Mechanical automation and climate room integration possible (SeedAIxpert HTC, SeedAIxpert HT)
SeedAIxpert handles seeds from any crops, vegetables, ornamentals, herbs, medical plants, or natural plants. We digitally assess quality and germination features of seeds of any shape, color or size – from sub-millimeter to centimeter range.
Our cutting-edge analysis system combines industry-grade cameras and lights with advanced analytical software, incorporating machine learning algorithms that allow advanced classification of seed and seedling quality.
You can accelerate the testing process already with manually taken images in a bench-top system, but we offer automation for high-throughput systems, too. Hundreds of seed testing trays can be automatically imaged and analyzed in short time with our automated SeedAIxpert HTC and HT systems. SeedAIxpert HT can be fit into climate-controlled rooms so that you can set optimal germination conditions for the seeds.
Our digital seed test analyses seed germination on paper (top-of-paper method) and germination on or in other substrates. For between-paper-tests, image acquisition after paper unfolding is possible, too. The digital germination test can handle samples in various testing methods like they are described in the ISTA or AOSA manuals.
The system can work with seedling emergence tests, too. Here, the outgrow of seedlings from a substrate, usually soil, is monitored. The appearance of a plant at a position in the substrate-filled tray is classified as emerged seedling. Even very early stages can be detected due to high sensitivity of the system.
With LemnaTec analytical software we process images to recognise the seeds and the emerging parts with classical image processing or with machine-learning assisted systems. Machine learning tools serve to classify the seedlings in more detail, for instance by detecting roots, root hairs, and shoots separately from the seed.
Based on the classification by LemnaTec analytical software, many informative factors such as percentage of germination or emergence, time of germination and metrical parameters of the seeds and seedlings are determined. By having a broad range of metical factors, data quality is strongly improved. Metrics comprise area, length, width, colour, or geometrical factors of seeds as well as roots, shoots and leaves that emerge from seeds during the germination process. Classification and metrics are decisive to discriminate normal seeds and seedlings form abnormal ones, e.g. damaged, infested, or strange-growing individuals.
Coated and uncoated seeds of most plant species can be analyzed. Depending on species and on eventual treatment or coating, seeds differ in size, color and shape. The SeedAIxpert algorithms can be adapted to many different seed types. Classification and quantification of seed characteristics are based on predefined criteria (number, size, color, and shape).
Examples include and are not limited to:
- Percentage of seed germination
- Germination rates (in time series analysis)
- Seedling emergence
- Seed and seedling morphological parameters (e.g. size, roundness, root length)
- Seed and seedling color and color distribution
LemnaTec Product Matrix
LemnaTec offers imaging and image processing technologies for several applications that suit for low to high throughput in laboratory, greenhouse or outdoor environments. Technologies are adapted to various sample types and can be combined with automation. All imaging systems are combined with dedicated image processing software, including user-programmable functions and machine learning.
