Potato Crop Disease Augmentation Dataset
Description
The PotatoCropDiseaseAugmentationDataset is a comprehensive collection of images designed to aid in the development and evaluation of machine learning models for detecting and classifying diseases in potato crops. This dataset includes augmented images representing both healthy and diseased potato plants, encompassing various stages of disease progression. The dataset is structured to support tasks such as disease detection, crop health monitoring, and disease classification. Subfolders in the Dataset: The dataset is organized into two main levels of subfolders: Top-Level Subfolders: augmented: Contains the disease categories and their corresponding augmented images. common_scab cut dry_rot gangrene healthy violet_root_rot Inside Each Disease Folder: Each disease or health status subfolder contains images of potato plants with different augmentations applied, including: Color_Intensity_Modifications Cutout_Occlusion Gaussian_Blur Geometric_Transformations Noise_Addition Perspective_Transformation Random_Brightness_Contrast Random_Cropping Random_Hue Super_Resolution ####Raw_Data:: Shuvo Kumar Basak. (2025). Potato Crop Disease and Health Dataset [Data set]. Kaggle. https://doi.org/10.34740/KAGGLE/DSV/10503599 Note for Researchers Using the dataset This dataset was created by Shuvo Kumar Basak. If you use this dataset for your research or academic purposes, please ensure to cite this dataset appropriately. If you have published your research using this dataset, please share a link to your paper. Good Luck.
Files
Steps to reproduce
Data Augmentation Procedure: To enrich the dataset and make it suitable for robust machine learning models, various data augmentation techniques have been applied. This enhances the diversity of the images, allowing the models to generalize better and perform accurately on unseen data. The following augmentation techniques are applied to each disease or health condition: Geometric Transformations: Rotation: Rotate the images by random angles (e.g., 90°, 180°, etc.). Flipping: Flip images horizontally or vertically. Resizing: Resize the images to different sizes such as 128x128 and 256x256. Translation: Shift images by a random amount to simulate positional variations. Color and Intensity Modifications: Brightness Adjustment: Increase or decrease the brightness to simulate different lighting conditions. Contrast Adjustment: Modify contrast to simulate varying lighting and exposure conditions. Saturation Adjustment: Adjust the color saturation to account for changes in environmental factors. Grayscale Conversion: Convert images to grayscale for models that need to process non-color data. Noise Addition: Gaussian Noise: Add random noise to simulate imperfections in real-world images. Salt and Pepper Noise: Randomly add white and black noise to images to simulate defects. Cutout and Occlusion: Randomly "cut out" sections of the image to simulate occlusion or damage to part of the potato plant. Perspective Transformation: Apply perspective changes to simulate different viewpoints of the potato plants. Super Resolution: Upscale low-resolution images to simulate high-definition captures, which helps the model adapt to varying quality inputs. Random Brightness and Contrast: Apply random adjustments to the brightness and contrast to simulate variations in the lighting during image capture. Random Cropping: Crop sections of the image randomly, which encourages the model to focus on different parts of the plant. Random Hue Adjustment: Modify the hue of the images to simulate changes in lighting or plant age. Gaussian Blur: Apply a Gaussian blur to the image to simulate blurry images due to poor focus or weather conditions. Each of these augmentations ensures that the model can learn to recognize diseases in various conditions, leading to better performance in real-world scenarios.