Varjovi, Mahdi HatamiAltun, SaraTalu, Muhammed FatihYeroglu, Celaleddin2024-08-042024-08-042018978-1-5386-6878-8https://hdl.handle.net/11616/98702International Conference on Artificial Intelligence and Data Processing (IDAP) -- SEP 28-30, 2018 -- Inonu Univ, Malatya, TURKEYSegmentation of tree images are used in agricultural applications such as harvest estimation. The accuracy of the segmentation process influences the success of such applications. Many methods are used for image segmentation such as thresholding, clustering, edge-based, region-based methods. The region growing algorithm is a very robust and easy to implement method, but the disadvantage of this method is that the threshold value is sensitive to critical and environmental noise. In this study, it is aimed to increase the segmentation quality of tree images. In tree segmentation; the optimum threshold values differ because of the camera's characteristics, the amount of light, the color of the leaf, the type of the fruity, the shadow of the branches and the other greens on the background. The starting point is automatically optimized with the help of genetic algorithms for the threshold values used in the determined region growing method.trinfo:eu-repo/semantics/closedAccesssegmentationthresholdregion growinggenetic algorithmConference Object2-s2.0-85062546576N/AWOS:000458717400100N/A