LEDCOM: A Novel and Efficient LED Based Communication for Precision Agriculture

Abstract

Wireless Sensor Networks and Satellite Remote Sensing are some of the existing techniques that are used to collect, analyze and interpret data from the agricultural crop sites. However, there are certain limitations common to both of these techniques that are concerned with the latency and the resolution of the data collected. UAVs (Unmanned Aerial Vehicles) are becoming another alternative that has become integral nowadays due to its affordable and scalable nature while offering user friendly requirements and customizations. This proposes a novel and cost-effective technique (LEDCOM) that harnesses the capabilities of ground sensors and unmanned UAV while using computer vision methods to produce a qualitative data analysis system that describes the crop site under supervision. An UAV is assumed to collect the ground based sensor node data in the form of binary patterns on LED Arrays that is encoded in the image taken by a camera of a drone. Image processing techniques are used to identify and decode the LED sequences from the arrays. The performance of the proposed system is evaluated under different features and image resolutions within the same lighting conditions. A promising performance is observed for LED pattern identification from the challenging images taken from a height.