农作物病虫草害图像识别诊断大数据集

拥有48种农作物700种病害的3289078张图片，102种虫害100万张图片，及400多种虫害300多万张图片。这些图片经精细过滤处理后，用于训练。通过图像识别算法、数据增强和深度学习，基于庞大的农作物病虫草害图片库，训练出识别模型。此模型能让农业生产者通过手机拍照迅速识别作物病虫草害，即时获取精准信息、防治措施、方法及用药施肥建议，提升农业生产效率与精准度。This dataset comprises 3,289,078 images corresponding to 700 disease types affecting 48 crop species, 1,000,000 images for 102 pest species, and over 3,000,000 images covering more than 400 pest species. All images have undergone rigorous fine-grained filtering before being utilized for model training. A crop disease, pest and weed recognition model was trained using image recognition algorithms, data augmentation techniques and deep learning methods, based on this large-scale specialized image database. This model enables agricultural practitioners to quickly identify crop diseases, pests and weeds by capturing photos with mobile phones, and instantly acquire accurate information, prevention and control measures, operational guidance as well as medication and fertilization recommendations, thereby improving agricultural production efficiency and precision.

搜集汇总

背景与挑战

背景概述

该数据集是一个大规模的农作物病虫草害图像识别资源，涵盖48种农作物、700种病害和数百种虫害，总计包含超过700万张经过处理的图片。它通过深度学习和数据增强技术训练出高效的识别模型，使农业生产者能够通过手机拍照快速诊断病虫草害，并即时获取防治建议，从而显著提升农业生产的精准度和效率。