The timely and accurate detection of plant diseases plays a crucial role in preserving agricultural productivity and ensuring optimal yield. Machine learning (ML) techniques offer promising solutions to address this challenge, particularly through the application of deep learning algorithms which excel in image processing tasks. Recent advancements in precision agriculture have increasingly leveraged deep learning models to identify plant diseases with high accuracy. In proposed work explored the effectiveness of nine deep neural networks in detecting plant diseases using various methodologies. Transfer learning and deep feature extraction techniques were employed to tailor these models to the specific task at hand. It focuses on pre-trained deep models for feature extraction and subsequent fine-tuning. By utilizing support vector machine (SVM), extreme learning machine (ELM), and k-nearest neighbor (KNN) classifiers, the extracted features were grouped and analyzed. Real-world images of plant diseases and pests from Turkey were used as the dataset for evaluation. Performance metrics such as accuracy, sensitivity, specificity, and F1-score were employed to assess the effectiveness of the models. The findings indicates that deep feature extraction coupled with SVM/ELM classification outperformed transfer learning approaches. Furthermore, the accuracy scores were notably higher for the fc6 layers of AlexNet, VGG16, and VGG19 models compared to other layers.column.