Scholarlink

<p>Weed control remains one of the most challenging problems in modern agriculture, where traditional practices, such as manual labor and widespread chemical herbicide use, often prove inefficient, labor-intensive, and environmentally damaging. In response to the rising need for sustainable and cost-effective solutions, this paper proposes an Automation-Driven Weed Removal Recommendation Engine (WRRE<b>)</b> designed to support precision agriculture through intelligent weed management strategies. The proposed algorithm categorizes the methods for treating weed infestations in localized areas of a crop field based on the percentage of weed presence within the crop to minimize impact on the environment and reduce the cost of labor by integrating variations in conventional weed elimination methods. The WRRE framework automates the classification of crop fields based on weed density, using a binary weed mask machine learning technique to quantify infestation levels from crop images. These insights enable the algorithm to categorize the field into defined weed density classes and recommend targeted interventions from manual and robotic hoeing to mechanical weeding and selective herbicide application. By applying context-specific strategies only where necessary, the system significantly reduces labor, chemical usage, and operational costs, while preserving soil health and improving crop yields. The experimentation for WRRE is done both in ideal scenario (WRRE -I) when all environmental conditions are favourable, and in real world scenario (WRRE -II) considering the factors such as crop type, weather conditions and soil moisture content that affect the weeding techniques. The model is implemented using Python within the Visual Studio Code environment and evaluated against traditional weeding practices. Experimental results highlight the system’s effectiveness in optimizing resource utilization, reducing environmental impact, and supporting scalable, automation-enabled weed control for modern agricultural ecosystems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Automated weed management in agriculture: the weed removal recommendation engine (WRRE)

  • 期刊论文

  • 发表日期：

    2026年5月14日

查看全文
  • Ritu Singhal,
  • Bhavika Srivastava,
  • Ishita Singhal,
  • Ananya Singhal

摘要

Weed control remains one of the most challenging problems in modern agriculture, where traditional practices, such as manual labor and widespread chemical herbicide use, often prove inefficient, labor-intensive, and environmentally damaging. In response to the rising need for sustainable and cost-effective solutions, this paper proposes an Automation-Driven Weed Removal Recommendation Engine (WRRE) designed to support precision agriculture through intelligent weed management strategies. The proposed algorithm categorizes the methods for treating weed infestations in localized areas of a crop field based on the percentage of weed presence within the crop to minimize impact on the environment and reduce the cost of labor by integrating variations in conventional weed elimination methods. The WRRE framework automates the classification of crop fields based on weed density, using a binary weed mask machine learning technique to quantify infestation levels from crop images. These insights enable the algorithm to categorize the field into defined weed density classes and recommend targeted interventions from manual and robotic hoeing to mechanical weeding and selective herbicide application. By applying context-specific strategies only where necessary, the system significantly reduces labor, chemical usage, and operational costs, while preserving soil health and improving crop yields. The experimentation for WRRE is done both in ideal scenario (WRRE -I) when all environmental conditions are favourable, and in real world scenario (WRRE -II) considering the factors such as crop type, weather conditions and soil moisture content that affect the weeding techniques. The model is implemented using Python within the Visual Studio Code environment and evaluated against traditional weeding practices. Experimental results highlight the system’s effectiveness in optimizing resource utilization, reducing environmental impact, and supporting scalable, automation-enabled weed control for modern agricultural ecosystems.