Çay bitkisindeki hastalıkların sinir ağları kullanılarak tespiti

This thesis investigates the automated detection of leaf illnesses in tea plants by combining optimization algorithms with cutting-edge computer methods, particularly neural networks. The need for agriculture is growing to maintain crop health with guarantee maximum yields, it is critical to promptly and accurately diagnose plant diseases. With Convolutional Neural Networks (CNNs) serving as the main architecture, the aim of this thesis is to build a reliable and effective technique that can identify several diseases that harm tea plants. The methodology entails the procurement from differing dataset of tea plant leaves, comprising the healthy and unhealthy specimens. Preprocessing the dataset takes into account variables including disease severity, illumination, and image resolution in order to improve model performance. The most recent optimization techniques, such as the mayfly and pelican optimization algorithms (MA), are used in this study to train the neural network using this well selected dataset. The method's performance in disease detection is estimated using metrics like accuracy, MSE, F-score and recall, and sensitivity. The CNN-POAMA model, which was proposed, attained values of 94.5%, 0.035, 0.91, 0.93, and 0.92, respectively. The findings of this study have important ramifications for the farming sector as well as for the growth of automated technologies that have the possible to totally convert the way tea plants manage illness. The model's scalability and potential for real-world application demonstrate how well neural networks and optimization algorithms work together to solve challenging agricultural problems.