Pollination of Palm Trees Using Drons and Artificial Intelligence

Abstract

‫‪quadcopter,‬‬‫‪a‬‬ ‫‪using‬‬ ‫‪pollination‬‬ ‫‪palm‬‬ ‫‪date‬‬ ‫‪for‬‬ ‫‪system‬‬ ‫‪smart‬‬ ‫‪a‬‬ ‫‪develop‬‬ ‫‪to‬‬ ‫‪aims‬‬ ‫‪memorandum‬‬ ‫‪This‬‬ ‫‪between‬‬‫‪distinguishing‬‬ ‫‪visually‬‬ ‫‪of‬‬ ‫‪capable‬‬ ‫‪model‬‬ ‫‪intelligence‬‬ ‫‪artificial‬‬ ‫‪an‬‬ ‫‪by‬‬ ‫‪supported‬‬ ‫‪urgent‬‬‫‪the‬‬ ‫‪of‬‬ ‫‪light‬‬ ‫‪in‬‬ ‫‪comes‬‬ ‫‪initiative‬‬ ‫‪This‬‬ ‫‪pollen.‬‬ ‫‪palm‬‬ ‫‪date‬‬ ‫‪female‬‬ ‫‪unpollinated‬‬ ‫‪and‬‬ ‫‪pollinated‬‬ ‫‪productivity,‬‬‫‪date‬‬ ‫‪enhance‬‬ ‫‪and‬‬ ‫‪effort‬‬ ‫‪manual‬‬ ‫‪reduce‬‬ ‫‪that‬‬ ‫‪solutions‬‬ ‫‪innovative‬‬ ‫‪for‬‬ ‫‪need‬‬ ‫‪the‬‬‫‪of‬‬ ‫‪chapter‬‬ ‫‪first‬‬ ‫‪The‬‬ ‫‪expansion.‬‬ ‫‪agricultural‬‬ ‫‪extensive‬‬ ‫‪with‬‬ ‫‪areas‬‬ ‫‪desert‬‬ ‫‪in‬‬ ‫‪especially‬‬ ‫‪its‬‬‫‪studying‬‬ ‫‪by‬‬ ‫‪palm,‬‬ ‫‪date‬‬ ‫‪the‬‬ ‫‪to‬‬ ‫‪related‬‬ ‫‪aspects‬‬ ‫‪theoretical‬‬ ‫‪the‬‬ ‫‪addresses‬‬ ‫‪memorandum‬‬ ‫‪exploring‬‬‫‪and‬‬ ‫‪flowers,‬‬ ‫‪female‬‬ ‫‪and‬‬ ‫‪male‬‬ ‫‪between‬‬ ‫‪distinguishing‬‬ ‫‪structure,‬‬ ‫‪morphological‬‬ ‫‪concept‬‬‫‪The‬‬ ‫‪methods.‬‬ ‫‪pollination‬‬ ‫)‪machine‬‬ ‫‪or‬‬ ‫‪(human‬‬ ‫‪artificial‬‬ ‫‪and‬‬ ‫)‪insect‬‬ ‫‪and‬‬ ‫‪(wind‬‬ ‫‪natural‬‬ ‫‪also‬‬‫‪were‬‬ ‫‪disadvantages‬‬ ‫‪and‬‬ ‫‪advantages‬‬ ‫‪their‬‬ ‫‪and‬‬ ‫‪technology,‬‬ ‫‪operating‬‬ ‫‪types,‬‬ ‫‪their‬‬ ‫‪drones,‬‬ ‫‪of‬‬ ‫‪and‬‬‫‪components‬‬ ‫‪its‬‬ ‫‪explaining‬‬ ‫‪project,‬‬ ‫‪this‬‬ ‫‪in‬‬ ‫‪used‬‬ ‫‪quadcopter‬‬ ‫‪the‬‬ ‫‪on‬‬ ‫‪focus‬‬ ‫‪a‬‬ ‫‪with‬‬ ‫‪addressed,‬‬ ‫‪unpollinated‬‬‫‪and‬‬ ‫‪pollinated‬‬ ‫‪of‬‬ ‫‪database‬‬ ‫‪field‬‬ ‫‪a‬‬ ‫‪chapter,‬‬ ‫‪practical‬‬ ‫‪the‬‬ ‫‪In‬‬ ‫‪mechanism.‬‬ ‫‪movement‬‬ ‫‪The‬‬‫‪platform.‬‬ ‫‪Roboflow‬‬ ‫‪the‬‬ ‫‪using‬‬ ‫‪classified‬‬ ‫‪manually‬‬ ‫‪and‬‬ ‫‪prepared‬‬ ‫‪was‬‬ ‫‪images‬‬ ‫‪pollen‬‬ ‫‪work‬‬‫‪The‬‬ ‫‪environment.‬‬ ‫‪Colab‬‬ ‫‪Google‬‬ ‫‪the‬‬ ‫‪in‬‬ ‫‪data‬‬ ‫‪this‬‬ ‫‪on‬‬ ‫‪trained‬‬ ‫‪then‬‬ ‫‪was‬‬ ‫‪model‬‬ ‫‪YOLOv8‬‬ ‫‪the‬‬‫‪training‬‬ ‫‪environment,‬‬ ‫‪software‬‬ ‫‪the‬‬ ‫‪up‬‬ ‫‪setting‬‬ ‫‪augmentation,‬‬ ‫‪preparation,‬‬ ‫‪data‬‬ ‫‪included‬‬model, and testing it on new data. The results showed that the model was able to classify images with high accuracy, despite challenges related to lighting, backgrounds, and class imbalance. A proposal was also made.