Apple Disease Dataset
Description
The dataset used in this study consists of images of Manalagi apples displaying a variety of visual conditions, both healthy and infected by diseases or damaged by pests. The data were collected from several apple orchards owned by local farmers in the Poncokusumo area, Malang Regency, Indonesia. Due to the high incidence of apple diseases in this region, discussions were conducted with farmers to identify the most common symptoms and types of damage affecting the harvest. This dataset focuses on six visual categories of apples, including the healthy fruit class and five classes representing major diseases and pest damage commonly found in the field. All images were captured in different orchards across Poncokusumo using a combination of digital SLR cameras and smartphones, including Realme 9 Pro (16 MP main camera), Apple iPhone 13 (12 MP dual camera), and Samsung Galaxy S Series (64 MP main camera). The data collection process was carried out with the assistance of agricultural experts to ensure accuracy and validity.
Files
Steps to reproduce
Data Collection: Images of Manalagi apples were collected from several orchards located in the Poncokusumo area, Malang Regency, Indonesia. The dataset includes both healthy apples and those affected by common diseases and pest damage. Field Verification: Discussions with local farmers and agricultural experts were conducted to identify the most common symptoms and disease types observed in apple plants. This ensured that the dataset accurately represents real field conditions. Image Acquisition: Photos were captured using various imaging devices, including a DSLR camera and several smartphones (Realme 9 Pro – 16 MP main camera, Apple iPhone 13 – 12 MP dual camera, and Samsung Galaxy S Series – 64 MP main camera). All images were taken under natural lighting conditions in the orchard environment. Data Validation and Categorization: The collected images were reviewed with the assistance of agricultural experts to confirm accuracy. Each image was then categorized into one of six classes (including the healthy category) based on visible disease or pest symptoms. Data Preparation: The final dataset was organized into separate folders for each category. Images were saved in JPG format with consistent resolution and labeled appropriately for use in machine learning and computer vision tasks.