使用优化混合U-Nets模型进行代细分和分类,以加强作物疾病诊断,使用优化混合U-Nets模型
Malathi Chilakalapudi1, Sheela Jayachandran1
1SCOPE, VIT-AP University, Amaravathi, Andhra Pradesh, India.
PeerJ. Computer science
|June 26, 2025
概括
这项研究引入了用于精确检测和分类农作物疾病的先进框架,大大提高了准确性并缩短了响应时间,以改善农业管理.
科学领域:
- 农业科学 农业科学
- 计算机视觉 计算机视觉
- 机器学习 机器学习
背景情况:
- 目前的作物疾病诊断方法缺乏精度,准确性和速度,阻碍了有效的管理.
- 现有的技术在分类准确性和及时检测方面存在局限性,影响作物产量.
研究的目的:
- 通过使用多面分析,开发一个改进的作物疾病检测和分类框架.
- 提高农业疾病诊断的精度,准确性和效率.
主要方法:
- 实施了适应性异型扩散用于农业图像无色化,以确保数据质量.
- 使用了 Fuzzy U-Net++ 模型,以增强图像细分和 fuzzy 决策.
- 介绍了移动大猩猩雷莫拉算法 (MGRA) 与卷积运算以实现最佳特征选择.
- 采用一个灵感来自LeNet的架构来进行疾病分类.
主要成果:
- 在疾病分类精度上取得了8.5%的改进,准确度提高了8.3%.
- 在召回方面表现出9.4%的改善,时间延迟减少4.5%.
- 曲线下的面积 (AUC) 增加了5.9%,特异性增加了6.5%.
结论:
- 与现有的方法相比,拟议的框架显著提高了作物疾病的检测和分类.
- 这一进步承诺通过精确性,准确性和及时性来实现更有效和高效的作物管理.
- 这项研究为农业健康的预防性措施铺平了道路,提高了作物弹性和产量.
相关概念视频
01:17Classification of Illness
8.0K
The meaning of illness is individualized to each person who experiences an alteration in health. In contrast, disease is a medical term indicating a pathological change in the structure and function of the body or mind. It is a condition that has specific symptoms and boundaries.
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
8.0K
01:26Classification of Systems-I
319
Linearity is a system property characterized by a direct input-output relationship, combining homogeneity and additivity.
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
319