In the future, agriculture will face the need for increasing production, sustainability, wisdom, and efficiency, which will bring significant challenges to the development of modern agriculture. With the gradual popularization of 5G, advanced information technologies such as the Internet of Things and artificial intelligence promoted the evolution of modern agriculture to intelligent agriculture. The 5G-based Internet of Things will play an essential role in the development of smart agriculture. This paper investigates the research progress of 5G Internet of Things in smart agriculture. It sorts out the development status of 5G smart agriculture Internet of Things in recent years. Following that, the concept of 5G smart agriculture Internet of Things is put forward. It expounds on the connotation, architecture, and enabling key technologies. According to the key application scenarios of smart agriculture, practical cases are presented, the development trend and application value of 5G smart agriculture Internet of Things are shown, and the future development direction is put forward. Firstly, the concept of smart agriculture is distinguished, and the category scenarios of smart agriculture are summarized. Following that, the current review research on 5G-IoT is analyzed. This paper focuses on the analysis and summary of the changes brought by 5G to various key scenarios in smart agriculture. This paper analyzes the related key technologies and challenges, puts forward some key scientific problems, and summarizes the research ideas. Finally, the development trend and application value of 5G smart agriculture Internet of Things are shown. The future development direction is also proposed.

Aiming at the current insulator defect detection algorithm can identify a single type of defects, and the accuracy is not high, a Dilated Re-parameterized Residual-YOLO(DRR-YOLO) algorithm is proposed, which is capable of identifying four defects of each of the four insulator types. Firstly, a new module DRR is proposed based on Dilated Re-param Block(DRB) and Dilation-wise Residual(DWR), which is combined with the C2f module of the YOLOv8 model to enhance the model’s detection effect for targets at different scales; Secondly, the Large Separable Kernel Attention SPPF(LSPPF) module is proposed to replace the SPPF in the original model, which enables the model to better preserve the feature information of the target while streamlining the structure; In addition, the advantages of MPDIoU and Inner-IoU are combined, and the loss function of the original model is replaced with Inner-MPDIoU. Finally, the DRR-YOLO model was compared with other existing models. Experiments demonstrate that the accuracy of the proposed model is superior to that of other models. Compared with the best - performing YOLOv5s and YOLOv8s, the mAP value on the validation set has increased by 1.9 percentage points and 2.2 percentage points respectively, reaching 94.7%. At the same time, it can be seen from the detection results that the proposed model has the highest confidence in detecting small targets.

WiFi fingerprint positioning is a common method for indoor location determination. Existing methods are susceptible to fluctuations in WiFi signal strength during the offline phase, leading to unevenly received signals. Additionally, during online positioning, there is a lack of integration with historical trajectory information. These problems can result in errors in both offline fingerprint acquisition and online location positioning. To address these problems, we propose a method that combines normality detection in the offline phase and Location Weighted K-nearest Neighbor positioning in the online phase. In the offline phase, initial Received Signal Strength Indication samples undergo preprocessing based on skewness and kurtosis for normality detection. If the samples conform to a normal distribution model, the probability density is estimated using the normal distribution function. If not, estimation occurs using the kernel density function model. Subsequently, values are averaged after Kalman filtering to establish a high-precision fingerprint database. During the online positioning phase, the LWKNN algorithm is employed. Initially, the Weighted K-nearest Neighbor method estimates the position, and this information is utilized as features to construct a Longterm and Shortterm Memory network model. The optimal path is determined through the least square method. Finally, the obtained outputs are integrated with historical data from the fingerprint positioning trajectory to enhance target positioning accuracy. Experimental results demonstrate that our indoor localization method significantly improves WiFi fingerprint localization accuracy compared to traditional localization methods.

The region of interest (ROI) with abundant and structured textures provides robust features in an indoor environment, which can effectively facilitate accurate simultaneous localization and mapping (SLAM). However, most existing visual SLAM systems generally treat ROI and non-ROI uniformly, resulting in ineffective employment of ROI. To meet this gap, we propose a robust saliency-driven visual SLAM system for indoor environments, coined RSD-SLAM. It can increase the focus on valuable ROI with the saliency maps obtained from a novel saliency prediction (SP) model. Specifically, we first design a saliency map construction method for visual SLAM enabling the SP model to accurately describe ROI, which generates the first indoor SP dataset integrating geometric, semantic, depth, and low-level visual information. Second, we develop a global stability constraint module for the SP model to enable the capability of keeping temporal consistency and illumination invariance. Third, we design a saliency map based hybrid saliency-driven mechanism to increase the focus of the system on ROI. At the front-end of the system, an adaptive feature-point extraction algorithm extracts more robust feature-points from ROI, and a saliency entropy based keyframe selection algorithm selects keyframes with the saliency value distribution of feature points. At the back-end, a dynamic weighted bundle adjustment (BA) optimization algorithm heavily weights the map points of ROI. Last, the particular focus on ROI results in a robust and accurate location. Extensive experiments, conducted on the EuRoC and TUM RGB-D datasets as well as in simulation environments, demonstrate that the proposed RSD-SLAM significantly outperforms the state-of-the-art in robustness and accuracy.