Camera-Based Vegetation Index from Unmanned Aerial Vehicles

Agriculture assumes a vital role in human life because it provides food, feed for livestock, and bioenergy. The agricultural sector is expected to meet the needs of secure and nutritious food for the community at all times to boost productivity. Providing nutrition, water and light precisely and measuredly is an important effort in plant cultivation to produce quality. This effort can be materialized by implementing smart farming involving devices and information technology. Vast field surveillance or monitoring is made easy with the advent of unmanned aerial vehicle (UAV). Detection of plant condition can be achieved by obtaining Vegetation Index (VI) through camera imaging in UAVs which are more economic compared to multispectral or hyperspectral cameras. This study aims to obtain VI that is accurate but still economical, so that it can be utilized even by small-scale agriculture. The work that will be done is to conduct repair experiments at several stages of image processing to produce a new, more accurate VI. The research stages started from experiments on previous research, to finding new research opportunities in VI. Furthermore, the experiment was carried out with the addition of white balance value parameters and other UAV sensor parameters at the Pre-Processing stage to improve its quality. The hypothesis of adding white balance parameters should prove to be more accurate in correcting shooting in various light conditions. Next, try to modify the feature extraction algorithm using Color Extraction Edge Detection. Followed by modifying it using Back Propagation Neural Network to increase accuracy at the image processing stage. After synthesizing some of these experiments, a new formula or model VI using the camera on the UAV is expected to be produced. This research will contribute to the modification of methods or algorithms at the image processing stage to produce a corrected image in producing a new VI that is more accurate using a camera on a more economical UAV.