视觉语言模型用于基于无人机的精准农业中零射击杂草检测和视觉推理
Muhammad Fahad Nasir1, Mobeen Ur Rehman2, Irfan Hussain1
1Khalifa University Center for Autonomous Robotic Systems, Khalifa University, Abu Dhabi, United Arab Emirates.
Frontiers in plant science
|February 16, 2026
概括
视觉语言模型 (VLMs) 显示了农业中精确杂草管理的前景. 双子闪 2.5 展示了强大的零拍摄性能和可解释性,为无人机图像分析提供了传统深度学习方法的低注释替代方案.
科学领域:
- 农业科学 农业科学
- 计算机视觉 计算机视觉
- 人工智能的人工智能
背景情况:
- 杂草显著降低排列作物的产量,需要有效的管理策略.
- 目前使用无人机图像检测杂草的深度学习方法需要大量的数据注释,并表现出不良的概括性.
- 现有模型的有限解释性阻碍了对农业应用的信任和采用.
研究的目的:
- 评估现代视觉语言模型 (VLMs) 在零射击环境中用于大豆田中杂草检测和管理的有效性.
- 评估VLM在识别杂草存在,空间定位,作物类型和生长阶段方面的能力.
- 引入和验证错误检测提示 (EPP) 以提高VLM的解释性和自我纠正.
主要方法:
- 六架商用VLM (ChatGPT-4.1,ChatGPT-4o,Gemini Flash 2.5,Gemini Flash Lite 2.5,LLaMA-4 Scout,LLaMA-4 Maverick) 使用来自大豆田的无人机图像进行了评估.
- 一个统一的提示程序旨在引起杂草的存在,空间定位,推理,作物生长阶段和作物类型.
- 错误检测提示 (EPP) 是一种反事实分析方法,用于测试模型的稳定性和自我纠正能力.
- 使用专家评级的基础性,特异性,可信性,非幻觉性和可操作性等分数量化解释性.
主要成果:
- 双子星闪 2.5 展示了最一致的零射击性能和最高的可解释性.
- 聊天GPT-4.1表现出强大的推理,但检测准确度较低,而聊天GPT-4o提供了平衡的性能.
- LLaMA-4变种在局部化和特异性方面表现出局限性.
- 双子星Flash Lite 2.5是高效的,但在EPP压力测试中缺乏稳定性,表明易碎的推理.
- 可解释性得分与空间正确性有积极的相关性,如视觉接地和文本到区域重叠指标所示.
结论:
- 视觉语言模型为精确的杂草管理提供了一个有希望的低注释方法.
- 在现场部署中,模型可靠性通过可解释性和适应性而不是仅仅通过规模更好地预测.
- 双子星Flash 2.5成为了表现最好的人,突出了VLM在实际农业应用中的潜力.
- 进一步研究VLM可解释性和反驱动的适应性对于推进自动化农业系统至关重要.
相关概念视频
01:15Depth Perception and Spatial Vision
2.1K
Depth perception is the ability to perceive objects three-dimensionally. It relies on two types of cues: binocular and monocular. Binocular cues depend on the combination of images from both eyes and how the eyes work together. Since the eyes are in slightly different positions, each eye captures a slightly different image. This disparity between images, known as binocular disparity, helps the brain interpret depth. When the brain compares these images, it determines the distance to an object.
2.1K
02:16Light Acquisition
9.7K
In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.
9.7K
01:24Vision
60.4K
Vision is the result of light being detected and transduced into neural signals by the retina of the eye. This information is then further analyzed and interpreted by the brain. First, light enters the front of the eye and is focused by the cornea and lens onto the retina—a thin sheet of neural tissue lining the back of the eye. Because of refraction through the convex lens of the eye, images are projected onto the retina upside-down and reversed.
60.4K
01:29Application of Linearization and Approximation
105
A drone flying through complex terrain often relies on more than one sensing method to estimate small changes in altitude. Along with direct measurements, air pressure provides a useful indirect indicator of vertical movement. Atmospheric pressure decreases as altitude increases, and this relationship is commonly described using an exponential model. Although accurate, converting pressure measurements into altitude values requires calculations that are too complex to perform repeatedly during...
105