K-Kotagiri (NRB) Avocado Leaf DataSet

The Avocado Image Dataset consists of 435 raw images showcasing both healthy and disease-infected avocados, with 216 images depicting avocados affected by various diseases and 219 images representing healthy specimens. Prior to annotation, the dataset underwent a splitting process with a ratio of 7:2:1 for training, validation, and testing respectively. Post-processing, all images were resized to 224x224 pixels. Annotation techniques, including vertical flipping, 90-degree rotation, and shearing, were exclusively applied during the training phase to enhance dataset diversity. The augmented dataset distribution across training, validation, and testing sets is outlined, showcasing a balanced representation of both healthy and diseased avocado images.

牛油果图像数据集（Avocado Image Dataset）共包含435张原始图像，涵盖健康与病害侵染两类牛油果样本，其中216张为受多种病害影响的染病牛油果图像，219张为健康牛油果样本图像。标注前，该数据集已按7:2:1的比例划分为训练集、验证集与测试集；后处理阶段中，所有图像均被统一调整至224×224像素尺寸。为提升数据集多样性，仅在训练阶段应用了包括垂直翻转、90度旋转与剪切变换在内的数据增强手段。增强后的数据集在训练集、验证集与测试集的分布情况已明确说明，两类样本在各子集内均保持均衡的占比。