Scholarlink

<p>A&#xa0;productive and environmentally friendly form of farming, hydroponics uses nutrient-rich water solutions to cultivate plants without the need for soil. In order to avoid crop loss, root infections must be accurately identified and detected early. The ensemble deep learning model (DL) proposed in this paper is specifically intended to identify plant root infections in hydroponic systems. Initially, the dataset is pre-processed by image Augmentation, Grayscale conversion, and Contrast Enhancement. Using the pre-processed results, the hydroponic plant regions are segmented by employing an Adaptive Fusion of Fuzzy-C-Means Region-Growing Algorithm (FCMRG), and the noisy regions are eliminated using 2‑level denoising by employing Gaussian and Median filters. The proposed DL model uses the segmented output to evaluate whether the plants are infected with root disease. Ultimately, the root conditions are detected using the Crayfish Ensemble Faster Region-based Convolutional Neural Network (CE-Faster RCNN). By using the PYTHON platform, the suggested technique attained an Accuracy of 99.34%. Thus, from the results, it is demonstrated that the suggested model performs superior to the other approaches. The implementation and source code are available at: <a href="https://github.com/mamatha67pp-hash/Root-Disease-Detection">https://github.com/mamatha67pp-hash/Root-Disease-Detection</a>.</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

Optimized Root Net: a Hybrid Approach to Plant Roots Disease Detection for Hydroponics Using a Deep Learning Ensemble Model

  • 期刊论文

  • 发表日期：

    2025年12月9日

查看全文
  • Mamatha V,
  • Ashwin Kumar UM,
  • Kavitha JC

摘要

A productive and environmentally friendly form of farming, hydroponics uses nutrient-rich water solutions to cultivate plants without the need for soil. In order to avoid crop loss, root infections must be accurately identified and detected early. The ensemble deep learning model (DL) proposed in this paper is specifically intended to identify plant root infections in hydroponic systems. Initially, the dataset is pre-processed by image Augmentation, Grayscale conversion, and Contrast Enhancement. Using the pre-processed results, the hydroponic plant regions are segmented by employing an Adaptive Fusion of Fuzzy-C-Means Region-Growing Algorithm (FCMRG), and the noisy regions are eliminated using 2‑level denoising by employing Gaussian and Median filters. The proposed DL model uses the segmented output to evaluate whether the plants are infected with root disease. Ultimately, the root conditions are detected using the Crayfish Ensemble Faster Region-based Convolutional Neural Network (CE-Faster RCNN). By using the PYTHON platform, the suggested technique attained an Accuracy of 99.34%. Thus, from the results, it is demonstrated that the suggested model performs superior to the other approaches. The implementation and source code are available at: https://github.com/mamatha67pp-hash/Root-Disease-Detection.