Elsanta strawberry plant

Octinion

Octinion was curious in how realistic a plant can be visualized in Virtual Reality. We took apon this challenge and made a 3D viewer where every aspect of the plant is viualized in great detail. Octinion’s strawberry robot picker determines with a visual algorithm which strawberry the robot can pick and which ones to avoid. The visual algorithm depends on machine learning to work. The more images the robot gets ,the more accurate the robot can pick strawberries. Therefore the intention and next steps is to procedurally generate the Elsanta plant so the robot can feed the machine learning algorithm with digital frames of the plant in every stage of its life. To do this manually, in the physical world, it would take years. VR lets us do this in a matter of weeks.

"Machine learning needs details"
"Creating the perfect strawberry"
"Translucency in leaves"

Project highlights

Details matter

To be able to interprete an image as a machine learning algorithm, the video stream created from this virtual world needed to be detailed enough. Otherwise the robot could interprete the data wrong and this would influence the learning of the robot.

Machine learning needs details

Image scanning

To make a strawberry feel real, image scanning was used to produce large textures. After layering a decent amount of textures the strawberry came to life.

Creating the perfect strawberry

Backlighting and translucency

The leaves where made out of the same principles as the strawberry. One more difficulty was to create translucency through the leaves. A shader was written to create this effect with realtime lighting. You can check out the effect in the video below.

Translucency in leaves

PROJECT SUMMARY

After the initial showcase of the 3D viewer, Octinion was convinced of the fact that these video stream images could be used to learn their algorithm much more efficiently.

We check our emails every day and reply even if we're at dinner.

LET'S WORK TOGETHER

Copied email to clipboard