Garden Powdery Mildew Identification Image Dataset

In current agricultural production, pests and diseases are one of the important factors affecting crop yield and quality, especially powdery mildew, which has a high incidence in various plants. Traditional manual detection methods are inefficient, easily affected by human factors, and cannot meet the needs of large-scale farm real-time monitoring. Existing machine learning algorithms lack accuracy in detecting powdery mildew and cannot effectively distinguish disease characteristics at different stages. This dataset aims to provide high-quality image samples for agricultural intelligence, with a focus on improving the precision and speed of powdery mildew detection.Data collection was done using high-resolution cameras, selecting samples from multiple plants collected under different times and lighting conditions, ensuring diversity and representativeness of the samples. Quality control of the data is ensured through multiple rounds of labeling and consistency checks, with the labeling team consisting of agricultural experts with a pathological background, totaling over 20 people. Data preprocessing includes image enhancement, noise reduction, and color correction steps to ensure the stability and accuracy of the data. All data is stored in JPG format, organized in a two-tier directory structure based on plant species and disease location.The core advantages of the dataset are manifested in several aspects. Firstly, the labeling accuracy is above 95%, and the labeling consistency assured by expert review exceeds 90%. In terms of technological innovation, new image enhancement techniques have been introduced to improve model robustness under different lighting conditions. The application value lies in being able to improve the accuracy of powdery mildew detection models by 15%, significantly enhancing the performance of agricultural disease automatic detection systems. Compared to similar datasets, this dataset has outstanding advantages in diversity and labeling accuracy and covers rare early symptoms of powdery mildew, making it extremely scarce. The dataset is well structured to support the expansion of different diseases and general applications.