In response to the current key issues in the field of smart irrigation for farmland, such as the lack of data sources and insufficient integration, a low degree of automation in drive execution and control, and over-reliance on cloud platforms for analyzing and calculating decision making processes, we have developed nodes and gateways for smart irrigation. These developments are based on the EC-IOT edge computing IoT architecture and long range radio (LoRa) communication technology, utilizing STM32 MCU, WH-101-L low-power LoRa modules, 4G modules, high-precision GPS, and other devices. An edge computing analysis and decision model for smart irrigation in farmland has been established by collecting the soil moisture and real-time meteorological information in farmland in a distributed manner, as well as integrating crop growth period and soil properties of field plots. Additionally, a mobile mini-program has been developed using WeChat Developer Tools that interacts with the cloud via the message queuing telemetry transport (MQTT) protocol to realize data visualization on the mobile and web sides and remote precise irrigation control of solenoid valves. The results of the system wireless communication tests indicate that the LoRa-based sensor network has stable data transmission with a maximum communication distance of up to 4 km. At lower communication rates, the signal-to-noise ratio (SNR) and received signal strength indication (RSSI) values measured at long distances are relatively higher, indicating better communication signal quality, but they take longer to transmit. It takes 6 s to transmit 100 bytes at the lowest rate of 0.268 kbps to a distance of 4 km, whereas, at 10.937 kbps, it only takes 0.9 s. The results of field irrigation trials during the wheat grain filling stage have demonstrated that the irrigation amount determined based on the irrigation algorithm can maintain the soil moisture content after irrigation within the suitable range for wheat growth and above 90% of the upper limit of the suitable range, thereby achieving a satisfactory irrigation effect. Notably, the water content in the 40 cm soil layer has the strongest correlation with changes in crop evapotranspiration, and the highest temperature is the most critical factor influencing the water requirements of wheat during the grain-filling period in the test area.

In view of the demand for a low-cost, high-throughput method for the continuous acquisition of crop growth information, this study describes a crop-growth monitoring system which uses an unmanned aerial vehicle (UAV) as an operating platform. The system is capable of real-time online acquisition of various major indexes, e.g., the normalized difference vegetation index (NDVI) of the crop canopy, ratio vegetation index (RVI), leaf nitrogen accumulation (LNA), leaf area index (LAI), and leaf dry weight (LDW). By carrying out three-dimensional numerical simulations based on computational fluid dynamics, spatial distributions were obtained for the UAV down-wash flow fields on the surface of the crop canopy. Based on the flow-field characteristics and geometrical dimensions, a UAV-borne crop-growth sensor was designed. Our field experiments show that the monitoring system has good dynamic stability and measurement accuracy over the range of operating altitudes of the sensor. The linear fitting determination coefficients (R²) for the output RVI value with respect to LNA, LAI, and LDW are 0.63, 0.69, and 0.66, respectively, and the Root-mean-square errors (RMSEs) are 1.42, 1.02 and 3.09, respectively. The equivalent figures for the output NDVI value are 0.60, 0.65, and 0.62 (LNA, LAI, and LDW, respectively) and the RMSEs are 1.44, 1.01 and 3.01, respectively.

To meet the demand of intelligent irrigation for accurate moisture sensing in the soil vertical profile, a soil profile moisture sensor was designed based on the principle of high-frequency capacitance. The sensor consists of five groups of sensing probes, a data processor, and some accessory components. Low-resistivity copper rings were used as components of the sensing probes. Composable simulation of the sensor’s sensing probes was carried out using a high-frequency structure simulator. According to the effective radiation range of electric field intensity, width and spacing of copper ring were set to 30 mm and 40 mm, respectively. A parallel resonance circuit of voltage-controlled oscillator and high-frequency inductance-capacitance (LC) was designed for signal frequency division and conditioning. A data processor was used to process moisture-related frequency signals for soil profile moisture sensing. The sensor was able to detect real-time soil moisture at the depths of 20, 30, and 50 cm and conduct online inversion of moisture in the soil layer between 0–100 cm. According to the calibration results, the degree of fitting (R2) between the sensor’s measuring frequency and the volumetric moisture content of soil sample was 0.99 and the relative error of the sensor consistency test was 0–1.17%. Field tests in different loam soils showed that measured soil moisture from our sensor reproduced the observed soil moisture dynamic well, with an R2 of 0.96 and a root mean square error of 0.04. In a sensor accuracy test, the R2 between the measured value of the proposed sensor and that of the Diviner2000 portable soil moisture monitoring system was higher than 0.85, with a relative error smaller than 5%. The R2 between measured values and inversed soil moisture values for other soil layers were consistently higher than 0.8. According to calibration test and field test, this sensor, which features low cost, good operability, and high integration, is qualified for precise agricultural irrigation with stable performance and high accuracy.

Single-modal images carry limited information for features representation, and RGB images fail to detect grass weeds in wheat fields because of their similarity to wheat in shape. We propose a framework based on multi-modal information fusion for accurate detection of weeds in wheat fields in a natural environment, overcoming the limitation of single modality in weeds detection. Firstly, we recode the single-channel depth image into a new three-channel image like the structure of RGB image, which is suitable for feature extraction of convolutional neural network (CNN). Secondly, the multi-scale object detection is realized by fusing the feature maps output by different convolutional layers. The three-channel network structure is designed to take into account the independence of RGB and depth information, respectively, and the complementarity of multi-modal information, and the integrated learning is carried out by weight allocation at the decision level to realize the effective fusion of multi-modal information. The experimental results show that compared with the weed detection method based on RGB image, the accuracy of our method is significantly improved. Experiments with integrated learning shows that mean average precision (mAP) of 36.1% for grass weeds and 42.9% for broad-leaf weeds, and the overall detection precision, as indicated by intersection over ground truth (IoG), is 89.3%, with weights of RGB and depth images at α = 0.4 and β = 0.3. The results suggest that our methods can accurately detect the dominant species of weeds in wheat fields, and that multi-modal fusion can effectively improve object detection performance.