Indian Apple Fruit Image Dataset with Advanced Preprocessing for Healthy and Rot Detection

This dataset presents a curated and preprocessed image collection of Indian apple fruits, specifically designed for machine learning and deep learning applications in fruit disease detection. It includes a total of 8,415 raw images, divided into 5,715 images of healthy apples and 2,700 images exhibiting rot disease symptoms. These images were captured in varied natural lighting and background conditions to ensure dataset diversity and real-world relevance. To enhance the utility of this dataset for computer vision research, each raw image was resized to a uniform resolution of 256\u00d7256 pixels and then subjected to 15 distinct preprocessing techniques. These image processing operations are widely used in deep learning pipelines for improving feature extraction, enhancing contrast, suppressing noise, and generating robust training variants. The applied preprocessing methods include:Original resized imageGrayscale conversionBlack & white binary thresholdingCLAHE (Contrast Limited Adaptive Histogram Equalization)Gaussian blurMedian blurBilateral filteringAdaptive mean thresholdingAdaptive Gaussian thresholdingCanny edge detectionSobel edge detectionGamma correction (\u03b3 = 1.5)HSV color jitteringImage rotation (+30 degrees)Horizontal flippingEach raw image thus generates 15 preprocessed variants, resulting in a significantly expanded dataset suitable for training robust models under a wide variety of augmented scenarios. The preprocessed images retain the original class labels (\healthy\ and \rot\) and follow a consistent naming convention for easy identification and integration into supervised learning pipelines. This dataset aims to support the development of AI-driven tools in agriculture, food quality inspection, and post-harvest disease management, particularly for Indian apple varieties. It is ideal for research on image classification, segmentation, transfer learning, and explainable AI (XAI) in agricultural contexts. The availability of both raw and processed images enables flexible experimentation, benchmarking, and model generalization studies. Researchers, students, and practitioners can use this dataset to train and evaluate models for automatic apple fruit quality assessment and disease diagnosis using computer vision techniques.