Article: Optimisation of weed management by image segmentation in precision agriculture Journal: International Journal of Computational Vision and Robotics (IJCVR) 2026 Vol.16 No.1 pp.41 - 54 Abstract: Accurate weed detection remains crucial for ultra-localised control in robotic solutions, simulating manual weeding in agriculture. Although many studies have been conducted in the field of weed detection using machine learning, most have focused mainly on direct detection, which can present challenges in the face of weed diversity. In this study, we propose an integrated approach based on vegetation/soil segmentation, followed by discrimination between crops and weeds using an object detector. Segmentation models such as UNet, FPN, and LinkNet have been thoroughly trained to discriminate efficiently between vegetation and soil. The results obtained are promising, with the trained models being able to generate binary images (masks) with an accuracy (Jaccard and Dice similarity indices) of over 89%. In addition, the execution speed reached 217 frames per second (fps). The integration of the localisation results from the detection model with the segmented images provides a robust method for accurately determining the position of weeds in the agricultural context, opening up new prospects for automated, targeted weed control solutions. Inderscience Publishers - linking academia, business and industry through research

Title: Optimisation of weed management by image segmentation in precision agriculture

Authors: Mohammed Habib; Salma Sekhra; Adil Tannouche; Youssef Ounejjar

Addresses: Faculty of Sciences, Spectrometry, Materials and Archaeomaterials Laboratory (LASMAR), Moulay Ismail University, Meknes, Morocco ' Faculty of Sciences, Spectrometry, Materials and Archaeomaterials Laboratory (LASMAR), Moulay Ismail University, Meknes, Morocco ' Laboratoire de l'Ingénierie et de Technologies Appliquées (LITA), Ecole Supérieure de Technologie de Béni Mellal, Université Sultan Moulay Slimane, Morocco ' Faculty of Sciences, Spectrometry, Materials and Archaeomaterials Laboratory (LASMAR), Moulay Ismail University, Meknes, Morocco

Abstract: Accurate weed detection remains crucial for ultra-localised control in robotic solutions, simulating manual weeding in agriculture. Although many studies have been conducted in the field of weed detection using machine learning, most have focused mainly on direct detection, which can present challenges in the face of weed diversity. In this study, we propose an integrated approach based on vegetation/soil segmentation, followed by discrimination between crops and weeds using an object detector. Segmentation models such as UNet, FPN, and LinkNet have been thoroughly trained to discriminate efficiently between vegetation and soil. The results obtained are promising, with the trained models being able to generate binary images (masks) with an accuracy (Jaccard and Dice similarity indices) of over 89%. In addition, the execution speed reached 217 frames per second (fps). The integration of the localisation results from the detection model with the segmented images provides a robust method for accurately determining the position of weeds in the agricultural context, opening up new prospects for automated, targeted weed control solutions.

Keywords: images segmentation; computer vision; convolutional neural networks; agricultural images; weed detection; smart farming; precision farming; deep learning.

DOI: 10.1504/IJCVR.2026.150338

International Journal of Computational Vision and Robotics, 2026 Vol.16 No.1, pp.41 - 54

Received: 17 Feb 2023
Accepted: 19 Dec 2023
Published online: 10 Dec 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Optimisation of weed management by image segmentation in precision agriculture

Keep up-to-date