Sub-surface bruise detection in apples
While the apple rotates on a roller line it is imaged from above and from the side in the short-wave infrared (SWIR) band. A bruise under the skin, invisible in colour, shows as a dark region in the infrared; an apple whose mask area exceeds the limit is diverted to the side lane.
Bruise area and position are logged for every fruit; for fruit above the limit a reject signal goes to the PLC and the pusher moves it into the chute. The result is shown on the operator display as two images side by side.
- 01 Infeed
- 02 Acquire
- 03 Analyse
- 04 Decide
- 05 Reject
Apples arrive at the station rotating on the roller line.
- Bruise area
- —
- Infrared contrast
- —
Scenes are illustrative simulations; they are not customer references or performance guarantees.
Need
A fresh bruise does not change the skin colour; the damage is in the tissue under the skin and a colour camera cannot tell it from sound fruit. The bruise darkens days later, by which time the fruit has been packed.
Solution approach
In bruised tissue cell water is released and absorbs short-wave infrared light more strongly than sound tissue. In the infrared image taken under broadband light the bruise appears as a dark region; it is segmented as a deviation from the sound-fruit model and compared with the visible image of the same spot. Top and side views plus rotation cover the whole surface.
Process steps
- 01
Infeed
Apples arrive at the station rotating on the roller line.
- 02
Acquire
The top camera and then the side camera take the infrared image.
- 03
Analyse
Dark regions are segmented and compared with the visible image.
- 04
Decide
The mask area is compared with the limit.
- 05
Reject
A bruised apple is pushed into the chute; sound apples stay on the line.
Related applications
Let us assess your application together
We test your sample in our laboratory and report on resolution, optics and lighting selection.