ST-F, POMM, CEA, URCA, TECX

The demonstrator is setting up an image acquisition system in the Vranken-Pommery presses facilities. The data analysis from the Vranken-Pommery presses allows the neural network training based on the quality classification of the grapes. The use of semantic segmentation enables the detection of target classes. According to grape maturity and possible diseases, fast decision-making regarding grape quality will allow the best trial of the grape crates. The short and precise intervention are the key challenges to optimise pressing logistics and the quality of the grape must.

Beyond state-of-the-art developments and impacts

With the implementation of the proposed quality control solution, it will be possible to automate the complex and human-dependent assessment of the grape's quality, reducing the analysis biases.