The rapid growth of the world’s population has put significant pressure on agriculture to meet the increasing demand for food. In this context, agriculture faces multiple challenges, one of which is weed management. While herbicides have traditionally been used to control weed growth, their excessive and random use can lead to environmental pollution and herbicide resistance. To address these challenges, in the agricultural industry, deep learning models have become a possible tool for decision-making by using massive amounts of information collected from smart farm sensors. However, agriculture’s varied environments pose a challenge to testing and adopting new technology effectively. This study reviews recent advances in deep learning models and methods for detecting and classifying weeds to improve the sustainability of agricultural crops. The study compares performance metrics such as recall, accuracy, F1-Score, and precision, and highlights the adoption of novel techniques, such as attention mechanisms, single-stage detection models, and new lightweight models, which can enhance the model’s performance. The use of deep learning methods in weed detection and classification has shown great potential in improving crop yields and reducing adverse environmental impacts of agriculture. The reduction in herbicide use can prevent pollution of water, food, land, and the ecosystem and avoid the resistance of weeds to chemicals. This can help mitigate and adapt to climate change by minimizing agriculture’s environmental impact and improving the sustainability of the agricultural sector. In addition to discussing recent advances, this study also highlights the challenges faced in adopting new technology in agriculture and proposes novel techniques to enhance the performance of deep learning models. The study provides valuable insights into the latest advances and challenges in process systems engineering and technology for agricultural activities.

Portable devices play an essential role where edge computing is necessary and mobility is required (e.g., robots in agriculture within remote-sensing applications). With the increasing applications of deep neural networks (DNNs) and accelerators for edge devices, several methods and applications have been proposed for simultaneous crop and weed detection. Although preliminary studies have investigated the performance of inference time for semantic segmentation of crops and weeds in edge devices, performance degradation has not been evaluated in detail when the required optimization is applied to the model for operation in such edge devices. This paper investigates the relationship between model tuning hyperparameters to improve inference time and its effect on segmentation performance. The study was conducted using semantic segmentation model DeeplabV3 with a MobileNet backbone. Different datasets (Cityscapes, PASCAL and ADE20K) were analyzed for a transfer learning strategy. The results show that, when using a model hyperparameter depth multiplier (DM) of 0.5 and the TensorRT framework, segmentation performance mean intersection over union (mIOU) decreased by 14.7% compared to that of a DM of 1.0 and no TensorRT. However, inference time accelerated dramatically by a factor of 14.8. At an image resolution of $1296\times 966$, segmentation performance of 64% mIOU and inference of 5.9 frames per second (FPS) was achieved in Jetson Nano’s device. With an input image resolution of $513\times 513$, and hyperparameters output stride OS = 32 and DM = 0.5, an inference time of 0.04 s was achieved resulting in 25 FPS. The results presented in this paper provide a deeper insight into how the performance of the semantic segmentation model of crops and weeds degrades when optimization is applied to adapt the model to run on edge devices. Lastly, an application is described for the semantic segmentation of weeds embedded in the edge device (Jetson Nano) and integrated with the robotic orchard. The results show good spraying accuracy and feasibility of the method.

Precision irrigation and optimization of water use have become essential factors in agriculture because water is critical for crop growth. The proper management of an irrigation system should enable the farmer to use water efficiently to increase productivity, reduce production costs, and maximize the return on investment. Efficient water application techniques are essential prerequisites for sustainable agricultural development based on the conservation of water resources and preservation of the environment. In a previous work, an off-policy deep reinforcement learning model, Deep Q-Network, was implemented to optimize irrigation. The performance of the model was tested for tomato crop at a site in Portugal. In this paper, an on-policy model, Advantage Actor–Critic, is implemented to compare irrigation scheduling with Deep Q-Network for the same tomato crop. The results show that the on-policy model Advantage Actor–Critic reduced water consumption by 20% compared to Deep Q-Network with a slight change in the net reward. These models can be developed to be applied to other cultures with high production in Portugal, such as fruit, cereals, and wine, which also have large water requirements.

Within the scope of precision agriculture, many applications have been developed to support decision making and yield enhancement. Fruit detection has attracted considerable attention from researchers, and it can be used offline. In contrast, some applications, such as robot vision in orchards, require computer vision models to run on edge devices while performing inferences at high speed. In this area, most modern applications use an integrated graphics processing unit (GPU). In this work, we propose the use of a tensor processing unit (TPU) accelerator with a Raspberry Pi target device and the state-of-the-art, lightweight, and hardware-aware MobileDet detector model. Our contribution is the extension of the possibilities of using accelerators (the TPU) for edge devices in precision agriculture. The proposed method was evaluated using a novel dataset of peaches with three cultivars, which will be made available for further studies. The model achieved an average precision (AP) of 88.2% and a performance of 19.84 frames per second (FPS) at an image size of 640 × 480. The results obtained show that the TPU accelerator can be an excellent alternative for processing on the edge in precision agriculture.

A weed plant can be described as a plant that is unwanted at a specific location at a given time. Farmers have fought against the weed populations for as long as land has been used for food production. In conventional agriculture this weed control contributes a considerable amount to the overall cost of the produce. Automatic weed detection is one of the viable solutions for efficient reduction or exclusion of chemicals in crop production. Research studies have been focusing and combining modern approaches and proposed techniques which automatically analyze and evaluate segmented weed images. This study discusses and compares the weed control methods and gives special attention in describing the current research in automating the weed detection and control. Keywords: Detection, Weed, Agriculture 4.0, Computational vision, Robotics

Fruit diseases are a continuous hazard to farmers. By applying computer vision-based techniques, precision agriculture can support the farmers in the decision making for fruit disease control. Features extraction is an essential task for the computer vision pipeline. Nowadays, in general, feature extraction for fruit diseases are handcrafted. However, empirical results in different domains confirm that features learned by Convolutional neural networks (CNNs) provide significant improvements in accuracy over handcrafted features. CNNs have been applied in many computer vision tasks, replacing the hand-engineered models. In general, a large-scale image dataset is necessary for training a CNN. However, there are not many fruit disease images available to compose the dataset. We propose to train a tiny and efficient deep convolutional network developed to run in the mobile devices to classify healthy peach fruits and three peach diseases. Based on transfer learning techniques and data augmentation strategies, the proposed model achieves a Macroaverage F1-score of 0.96. The model does not misclassify any disease class. This achievement shows the potential of using small CNN models for fruit disease classification when having a small quantity of training data.