Deep Neural Networks to Detect Weeds from Crops in Agricultural Environments in Real-Time: A Review

Rakhmatuiln, Ildar; Kamilaris, Andreas; Andreasen, Christian

doi:10.3390/rs13214486

Cited by 60 publications

(35 citation statements)

References 148 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Advances in AI and field robotics makes it possible to precisely locate and identify weed and crop plants (Gikunda and Jouandeau, 2019;Rakhmatulin et al, 2021) and direct the laser beam toward the meristems of the weed seedlings for real-time laser control. Assuming single spot applications is used and there is limited heat movement beyond the irradiate spot, only a small proportion of the area in the field will be exposed to the treatment.…”

Section: Treated Areamentioning

confidence: 99%

“…(Vitali et al, 2021). As long as the recognitions system only separates the crop from other plants, it can be done with a high precision (93%) in realtime (Rakhmatulin et al, 2021). A complex recognition and decision system, that locate and identify many different weed species and decide which one needs to be irradiated may require more processing time than a simple system resulting in a slower driving speed.…”

Section: Efficacy Of the Lasermentioning

confidence: 99%

“…Identification of the target is essential to limit energy consumption. Recognition tools based on artificial intelligence make it possible to distinguish weeds from crop plants in agricultural environments in real-time (Wang et al, 2019;Rakhmatulin et al, 2021). The meristem can be detection by using high resolution cameras, while the positioning can be guided by precise laser scanners.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Laser Weeding With Small Autonomous Vehicles: Friends or Foes?

2022

Self Cite

View full text Add to dashboard Cite

Weed control is necessary to ensure a high crop yield with good quality. Herbicide application and mechanical weeding are the most common methods worldwide. The use of herbicides has led to the increasing occurrence of herbicide-resistant weeds and unwanted contamination of the environment. Mechanical weed control harms beneficial organisms, increases the degradation of organic matter, may dry out the soil, and stimulate new cohorts of weed seeds to germinate. Therefore, there is a need to develop more sustainable weed control means. We suggest using small autonomous vehicles equipped with lasers as a sustainable alternative method. Laser beams are based on electricity, which can be produced from non-fossil fuels. Deep learning methods can be used to locate and identify weed and crop plants for targeting and delivery of laser energy with robotic actuators. Given the targeted nature of laser beams, the area exposed for weed control can be reduced substantially compared to commonly used weed control methods. Therefore, the risk of affecting non-target organisms is minimized, and the soil will be kept untouched in the field, avoiding triggering weed seeds to germinate. Small autonomous vehicles may have limited weeding capacity, and precautions need to be taken as reflections from the laser beam can be harmful to humans and animals. In this paper, we discuss the pros and cons of replacing or supplementing common used weed control methods with laser weeding. The ability to use laser weeding technology is relatively new and not yet widely practiced or commercially available. Therefore, we do not discuss and compare the costs of the various methods at this early stage of the development of the technology.

show abstract

Section: Treated Areamentioning

confidence: 99%

Section: Efficacy Of the Lasermentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Laser Weeding With Small Autonomous Vehicles: Friends or Foes?

2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…Therefore, it is essential to deploy resources to monitor the growth of weeds and reduce weeds for healthy crop cultivation. There are two traditional strategies that are used to reduce the influence of weeds: mechanical weed control (e.g., mowing, mulching and tilling) and chemical weed control (i.e., using herbicides; Rakhmatulin et al, 2021). Both strategies have drawbacks.…”

Section: Introductionmentioning

confidence: 99%

Semi-supervised Learning for Weed and Crop Segmentation Using UAV Imagery

2022

View full text Add to dashboard Cite

Weed control has received great attention due to its significant influence on crop yield and food production. Accurate mapping of crop and weed is a prerequisite for the development of an automatic weed management system. In this paper, we propose a weed and crop segmentation method, SemiWeedNet, to accurately identify the weed with varying size in complex environment, where semi-supervised learning is employed to reduce the requirement of a large amount of labelled data. SemiWeedNet takes the labelled and unlabelled images into account when generating a unified semi-supervised architecture based on semantic segmentation model. A multiscale enhancement module is created by integrating the encoded feature with the selective kernel attention, to highlight the significant features of the weed and crop while alleviating the influence of complex background. To address the problem caused by the similarity and overlapping between crop and weed, an online hard example mining (OHEM) is introduced to refine the labelled data training. This forces the model to focus more on pixels that are not easily distinguished, and thus effectively improve the image segmentation. To further exploit the meaningful information of unlabelled data, consistency regularisation is introduced by maintaining the context consistency during training, making the representations robust to the varying environment. Comparative experiments are conducted on a publicly available dataset. The results show the SemiWeedNet outperforms the state-of-the-art methods, and its components have promising potential in improving segmentation.

show abstract

“…The application of ML models ranging from classical classification and regression algorithms, to state-of-the-art deep neural networks has played an important role in the advancement of remote and proximal sensing technologies, contributing to a wide range of domains. One such important applicative field is precision agriculture [11,15,16]. Indeed, advancements in ML techniques applied to computer vision have allowed the adoption of these proximal sensing technologies for the detection, identification, and measurement of plants and harvest [17].…”

Section: Introductionmentioning

confidence: 99%

Exploratory Analysis on Pixelwise Image Segmentation Metrics with an Application in Proximal Sensing

et al. 2022

View full text Add to dashboard Cite

A considerable number of metrics can be used to evaluate the performance of machine learning algorithms. While much work is dedicated to the study and improvement of data quality and models’ performance, much less research is focused on the study of these evaluation metrics, their intrinsic relationship, the interplay of the influence among the metrics, the models, the data, and the environments and conditions in which they are to be applied. While some works have been conducted on general machine learning tasks such as classification, fewer efforts have been dedicated to more complex problems such as object detection and image segmentation, in which the evaluation of performance can vary drastically depending on the objectives and domains of application. Working in an agricultural context, specifically on the problem of the automatic detection of plants in proximal sensing images, we studied twelve evaluation metrics that we used to evaluate three image segmentation models recently presented in the literature. After a unified presentation of these metrics, we carried out an exploratory analysis of their relationships using a correlation analysis, a clustering of variables, and two factorial analyses (namely principal component analysis and multiple factorial analysis). We distinguished three groups of highly linked metrics and, through visual inspection of the representative images of each group, identified the aspects of segmentation that each group evaluates. The aim of this exploratory analysis was to provide some clues to practitioners for understanding and choosing the metrics that are most relevant to their agricultural task.

show abstract

Deep Neural Networks to Detect Weeds from Crops in Agricultural Environments in Real-Time: A Review

Cited by 60 publications

References 148 publications

Laser Weeding With Small Autonomous Vehicles: Friends or Foes?

Laser Weeding With Small Autonomous Vehicles: Friends or Foes?

Semi-supervised Learning for Weed and Crop Segmentation Using UAV Imagery

Exploratory Analysis on Pixelwise Image Segmentation Metrics with an Application in Proximal Sensing

Contact Info

Product

Resources

About