Wireless Sensor Networks (WSN) is now widely used in precision agriculture applications. Sensors play an important role in WSN.The traditional humidity sensors employed in the agriculture have a drawback that they acquire a large amount of data which is to be transmitted or processed. Whereas the intelligent sensor proposed in this paper uses a conventional sensor and an embedded processor that can process the data acquired using algorithms to reduce the amount of data collected, to extract only relevant information and to present this information in a format which minimizes post-processing latency. As a result of which the final processing end has to perform a little computation that too only on features (like significant changes in humidity) rather on the huge data hence reducing the power consumption. The intelligent humidity sensor reduces the amount of data processed by 50% (depending upon humidity variations) and thereby also reducing the power consumption.