相关实验视频
Updated: Jul 30, 2026
08:47
Computer Vision-Based Biomass Estimation for Invasive Plants
Published on: February 9, 2024
基于GAN的合成作物疾病图像生成,用于精确的作物疾病识别
Chao Wang1,2,3, Yuting Xia1,2,3, Lunlong Xia1,2,3
1School of Information and Artificial Intelligence, Anhui Agricultural University, Hefei, 230036, China.
Plant methods
|March 30, 2025
概括
这项研究引入了一个新型网络 (FHWD) 来生成现实的作物疾病图像,解决深度学习模型的数据稀缺问题. 通过提高图像质量和细节,FHWD显著提高了疾病检测准确性和模型概括性.
科学领域:
- 农业技术 农业技术
- 计算机视觉 计算机视觉 计算机视觉
- 深度学习是一种深度学习.
背景情况:
- 用于检测作物疾病的深度学习模型需要广泛的,高质量的图像数据.
- 农作物疾病图像的有限可用性阻碍了模型的准确性和概括性.
研究的目的:
- 开发一种有效的图像增强技术,以解决作物疾病图像的稀缺问题.
- 提高深度学习模型的准确性和通用性,以识别作物疾病.
主要方法:
- 设计了一个频域和波形图像增强网络 (FHWD),具有双区分器结构.
- 利用波形损失和Fast Fourier转换损失函数,专注于图像细节和频率组件.
- 采用高频分辨器,以确保生成的图像纹理和结构中的真实性.
主要成果:
- 与其他模型相比,FHWD生成的图像具有更现实的叶病病变和优越的视觉质量.
- 使用FHWD的增强数据提高了对番茄叶病的VGG16,GoogleNet和ResNet18模型的分类准确度,平均为7.25%.
- 该网络有效地解决了数据稀缺问题,为疾病识别提供了更丰富的培训数据.
结论:
- FHWD是增强作物疾病图像数据集的强大工具,对于改进自动疾病监测至关重要.
- 拟议的方法提高了疾病病变和纹理的真实性,从而提高了深度学习模型的性能.
- 这种方法有助于通过改善数据可用性,使作物疾病检测系统更强大,更准确.
相关概念视频
00:53Transgenic Organisms
Overview
02:50Transgenic Plants
Recombinant DNA technology called transgenesis is often used to add a foreign gene or remove a detrimental gene from an organism. Such genetically modified organisms are called transgenic organisms.
The first-ever transgenic plant was a tobacco plant developed in 1983 that showed resistance against the tobacco mosaic virus. Since then, many transgenic plants have been developed and commercialized for improving the agricultural, ornamental, and horticultural value of a crop plant. Transgenic...
The first-ever transgenic plant was a tobacco plant developed in 1983 that showed resistance against the tobacco mosaic virus. Since then, many transgenic plants have been developed and commercialized for improving the agricultural, ornamental, and horticultural value of a crop plant. Transgenic...