Fecha de publicación: 11/02/2021 Fuente: WIPO Agriculture Portada According to a report of the Food and Agriculture Organization, the world population is projected to increase another two billion by 2050, while the arable region is expected to only grow by 5%. To increase productivity in agriculture, clever and effective agricultural techniques are therefore required. The determination of land adequacy for agriculture is a basic instrument for the growth of agriculture. Agriculture is using many emerging technologies and inventions as an approach to the gathering and distribution of agricultural knowledge. The rapid growth of wireless sensor networks has led to the creation of Internet of Things (IoT) of low cost and compact sensor devices, which are a feasible method for automating and making decisions in agriculture. This study proposes an expert framework for the evaluation of agricultural land suitability, by combining sensor networks with artificial intelligence systems such as neural networks and multi-component perceptron. This innovation would help farmers determine the farmland for agriculture according to four classes of judgment, i.e. more acceptable, suitable, reasonably appropriate and unsuitable. This measurement is based on the data from the different sensor instruments used for machine training. The results achieved by MLP with four hidden layers are seen in contrast to the other current model, to be successful for the multiclass classification system. This learned model is used to evaluate potential evaluations and to identify the land after each harvest.
Al-BASED CROP RECOMMENDATION SYSTEM FOR SMART FARMING
TOWARDS AGRICULTURE 5.0
Diagram:
Figure 1: Flow network for Al based crop management system