Seed counting and purity analysis
The sample is spread onto a light table by a vibratory feeder. Shape measures come from the backlit silhouette and colour from the image taken under diffuse top light; the seeds are counted one by one, classified, and the purity of the sample is reported.
Seed count, class distribution and purity are written to the laboratory record together with the sample number. If purity is below the limit the sample is rejected, and the result is shown on the operator display in class colours.
- 01 Spread
- 02 Silhouette
- 03 Colour
- 04 Count
- 05 Report
The feeder spreads the sample on the light table in a single layer.
- Seeds counted
- —
- Purity
- —
- Off-class seeds
- —
Scenes are illustrative simulations; they are not customer references or performance guarantees.
Need
Seeds are small and lie touching one another on the table, so two touching seeds can be counted as one blob. A broken seed and a small one, or an empty husk and a full seed, cannot be separated by colour alone, and seeds of a foreign species are often the same tone.
Solution approach
The backlight turns every seed into a sharp silhouette from which area, length, width and contour curvature are measured; touching seeds are split at the necks of the contour. An empty husk lets some light through, so its silhouette stays pale. The colour image taken under the ring light is then matched to the same seeds, and the class is decided from shape and colour together.
Process steps
- 01
Spread
The feeder spreads the sample on the light table in a single layer.
- 02
Silhouette
The backlight comes on and the silhouette image of the seeds is taken.
- 03
Colour
The ring light comes on and a colour image of the same seeds is taken.
- 04
Count
Touching seeds are separated; every seed is counted and classified.
- 05
Report
Purity is compared with the limit; the table is wiped clear.
Related applications
Let us assess your application together
We test your sample in our laboratory and report on resolution, optics and lighting selection.