Apple Disease Detection 800
Description
This dataset contains 800 annotated apple leaf and fruit images for apple disease detection using deep learning and computer vision techniques. The dataset is organized into train, validation, and test sets in YOLOv8 format. It was prepared for object detection and classification tasks related to identifying diseased and healthy apples. The images were preprocessed using auto-orientation and resized to 640×640 resolution. This dataset can be used for training, validating, and evaluating machine learning models for agricultural disease detection systems and smart farming applications. Dataset Split: Train Images: 560 Validation Images: 120 Test Images: 120 Total Images: 800
Files
Steps to reproduce
Download the dataset from Mendeley Data. Extract the ZIP file. Install the required Python libraries such as Ultralytics YOLOv8, OpenCV, NumPy, and PyTorch. Open the dataset folder and use the data.yaml file for YOLOv8 training configuration. Train the model using the YOLOv8 training command. Validate the trained model using the validation dataset. Test the model performance on the test dataset. Evaluate detection accuracy, precision, recall, and mAP metrics.
Institutions
- Khwaja Yunus Ali UniversityRajshahi Division, Chauhāli