2018年用于杂草叶子计数的图像数据集

叶子计数数据集的创建目的是通过计算杂草叶子数量来估计杂草的生长阶段，以优化除草剂喷洒方式并高效去除杂草。该数据集包含9372张png 格式和不同分辨率的红绿蓝图像，用于18种不同生长阶段的杂草。数据集根据叶子数量分为9类，即每株植物的数量，从1到9片以上不等，每个类别包含160到3292张图片。该数据集非常适合评估关于叶数的多类图像分类任务，但不适用于杂草种类分类和语义分割。https://vision.eng.au.dk/leaf-counting-dataset/

The Leaf Counting Dataset was developed to estimate the growth stage of weeds by counting their leaf numbers, so as to optimize herbicide spraying strategies and achieve efficient weed removal. This dataset contains 9,372 RGB images in PNG format with varying resolutions, covering weeds at 18 distinct growth stages. It is divided into 9 categories based on the number of leaves per plant, ranging from 1 to more than 9 leaves, with each category containing between 160 and 3,292 images. This dataset is highly suitable for evaluating multi-class image classification tasks related to leaf count, but is not applicable to weed species classification and semantic segmentation. https://vision.eng.au.dk/leaf-counting-dataset/