Received 10 January 2023, accepted 27 January 2023, date of publication 3 February 2023, date of current version 8 February 2023. Digital Object Identifier 10.1109/ACCESS.2023.3242604 Saliency-Based Semantic Weeds Detection and Classification Using UAV Multispectral Imaging ANUM NAVEED 1 , WASIF MUHAMMAD 1 , (Member, IEEE), MUHAMMAD JEHANZEB IRSHAD 1 , MUHAMMAD JAVAID ASLAM 1 , SAJJAD MANZOOR 2 , TASLEEM KAUSAR 2 , AND YUN LU 3 1 Intelligent Systems Laboratory, Department of Electrical Engineering, University of Gujrat, Gujrat 50400, Pakistan 2 Mirpur University of Science and Technology (MUST), Mirpur, Azad Jammu and Kashmir 10250, Pakistan 3 School of Computer Science and Engineering, Huizhou University, Huizhou, Guangdong 516007, China Corresponding author: Yun Lu (luyun_hit@163.com) This work was supported in part by the fund which aims to improve scientifc research capability of key construction disciplines in Guangdong province ‘‘Light-Weight Federal Learning Paradigm and Its Application’’ under Grant 2022ZDJS058, in part by the Professoral and Doctoral Scientifc Research Foundation of Huizhou University under Grant 2020JB058, and in part by the Research Project of Enhanced Independent Innovation Ability of Huizhou University under Grant hzu202018. ABSTRACT Weeds infestation causes damage to crops and limits the agricultural production. The traditional weeds controlling methods rely on agrochemicals which demand labour-intensive practices. Various methods are proposed for the pursuit of weeds detection using multispectral images. The machine vision-based weeds detection methods require the extraction of a large number of multispectral texture features which in turn increases the computational cost. Deep neural networks are used for pixel-based weeds classifcation, but a drawback of these deep neural network-based weeds detection methods is that they require a large size of images dataset for network training which is time-consuming and expensive to collect particularly for multispectral images. These methods also require a Graphics Processing Unit (GPU) based system because of having high computational cost. In this article, we propose a novel weeds detection model which addresses these issues, as it does not require any kind of supervised training using labelled images and multispectral texture features extraction. The proposed model can execute on a Central Processing Unit (CPU) based system as a result its computational cost reduces. The Predictive Coding/Biased Competition-Divisive Input Modulation (PC/BC-DIM) neural network is used to determine multispectral fused saliency map which is further used to predict salient crops and detect weeds. The proposed model has achieved 94.38% mean accuracy, 0.086 mean square error, and 0.291 root mean square error. INDEX TERMS Predictive coding biased competition divisive input modulation (PC/BC-DIM) network, saliency map, pixel-based weeds classifcation, robotic weeds detection, weeds infestation. I. INTRODUCTION The obnoxious growth of invasive weeds is a critical issue that needs to be controlled through an economically suitable method for maximum yield of crops on hectares of land. It is necessary to detect and classify weeds accurately so that the herbicide spray can be optimised with selective weeds treatment for a specifc region of weeds instead of spraying on hectares of land in order to reduce the manual effort of farm- ers. Traditional weeds removal methods include: mechan- ical, cultural, and chemical methods. The manual weeds The associate editor coordinating the review of this manuscript and approving it for publication was Hossein Rahmani . eradication method is economical and suitable for the small area of land but for hectares of land this method is not effcient. The cultural weeds controlling approach includes techniques such as crop rotation and using competitive forage species. Mechanical weeds removal involves methods such as tillage, mowing, and mulching. This method is slow and requires suitable weather and land conditions for its imple- mentation. Chemical weeds removal techniques have harm- ful impacts due to the utilization of herbicides which have a life-threatening infuence on aquatic and land organisms. Biological weeds removal methods such as pathogens and arthropods are used but these are expensive and fail because of its specifcity for a particular weeds. VOLUME 11, 2023 This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/ 11991