Artificial vision is becoming an essential in digital farming, for IoT connected sensors and for sensors embedded on tractors or robots. The use of low cost, low power consumption and high-performance neuromorphic circuits in artificial vision tasks aims to enforce innovation in smart agriculture, to reduce the costs of the vision-based tools, thus facilitating their adoption in farms. These technologies will help to address agricultural challenges such as reducing the use of pesticides or increasing the quality and the profitability of agricultural production.
Artificial Intelligence (AI) algorithms have been used to have informed expectation on which pests and diseases have more probabilities to happen each year based on the weather forecasted, soil conditions, etc. Been able to predict in advance these events will help farmers on the selection of the most pest and disease resistant variety to cultivate in the coming season to minimize their losses and to maximize yield. Neuromorphic platforms are expected to increase the accuracy of the prediction compared with the traditional platforms.
Within this application domain, these technologies will be demonstrated in the Use-case 2.1 Automatic weeding (essentially weed and crop detection and identification) and Use case 2.2 Tomato pests and diseases forecast leading to variety selection.