A four-arm harvesting robot system was designed to integrate with the fruit picking-collecting-transporting multifunction for the apples’ automatic harvesting. Taking the standardized tall-spindle and dwarf-rootstock apple tree as the object, the target operational area was determined for the harvesting robot, according to the fruits’ spatial distribution within the tree canopy. A new configuration of a four-arm picking manipulator and the operational mode were proposed with the four Cartesian coordinate arms in the three degree-of-freedom (DOF). An electric-pneumatic hybrid dual-stage driving structure was utilized to ensure efficient and large-scale telescopic motion within the tree canopy. Additionally, a CAN open bus-based integrated drive-control harvesting gripper was designed to enable efficient harvesting operations via a combination of fruit gripping and twisting actions. A multi-task deep convolutional network was adopted to recognize the fruit’s discrete visual pixel areas that were caused by branches and leaves occlusion. As such, the semantic segmentation of the occluded fruits and end-to-end determination of the discrete areas’ ownership were realized to overcome the traditional single-task networks in the classification of discrete regions of the same fruit. The view frustum projection model was introduced to locate the centroid of the target fruit, according to the local point cloud information on the surface. A novel strategy of four-arm picking task area partitioning was proposed, according to the clustered distribution characteristics of fruits within the tree canopy. The time-optimal four-arm collaborative picking task planning was also proposed to achieve the efficient traversal of different regions inside the tree canopy by the robotic arms. Finally, the key components of the harvesting robot were integrated to develop the autonomous harvesting workflow. The production trials were also conducted in a high-density dwarf rootstock orchard. The results showed that the recognition rate on the visible fruits was 92.94%, among which 90.27% of the fruits’ positioning accuracy was sufficient for picking operations. The robot’s average overall picking efficiency was 7.12 seconds per fruit, among which the efficiencies of single-, dual-, and four-arm were 9.59, 8.17, and 4.87 seconds per fruit, respectively. The efficiency of four-arm collaborative picking was approximately 1.96 times that of single-arm picking. The success harvesting rate of visible fruits was 82.00%, and the overall harvesting rate for all fruits inside the tree canopy was 74.56%. The success rate of harvesting reached up to 100% in the outer peripheral areas of the tree canopy where the fruits were sparse. However, the success rates of target recognition, location, and operation were significantly lower in the inner-dense region where the fruits were intensive, resulting in a harvesting success rate of 73.63%. The harvesting failures were attributed to the fruits that were obstructed by branches and leaves, leading to the visual recognition and positioning accuracy, as well as the interference and collisions with the harvesting manipulator. Therefore, the robot's capability of autonomous obstacle avoidance was enhanced to improve the tree structure and the performance of this harvesting robot. This finding can be considered as the preliminary exploration for the development and application of robotic harvesting models for freshly-eaten fruits.

Cucurbit grafting and seedling cultivation can often require the directional seeding in modern agriculture. But manual directional seeding of the cucurbit rootstocks grafting can suffer from the high labor intensity, the low operational efficiency and directional accuracy. It is an extremely urgent need for the highly efficient and precise directional seeding equipment in the seedling cultivation market. The conventional tray seeders are also limited to the directional seeds. In this study, a robotic system was designed for the precision directional seeding of the white pumpkin seeds using machine vision. A vibratory seed supply mechanism was integrated with an adsorption effector, a handling mechanism, and a machine vision system. The directional seeding robot was accurately identified the morphological features of the cucurbit rootstock seeds. Some information, such as the seed bud angle, was acquired using machine vision. The seed pickup and orientation actuator were then guided to pick up, reorient, and place the seeds. The synchronous directional sowing of five seeds per row was realized after following procedures. Firstly, an image processing with the OpenCV was developed to extract the seed contour, and then identify the critical features, including the bud position, geometric center, and bud point angle. Secondly, a suction end-effector was designed for the negative pressure adsorption and servo-driven rotation. A dynamic model was constructed for the internal flow field within the suction nozzle. The key parameters were then determined, such as the nozzle shape, diameter, and adsorption pressure. An electromagnetic vibratory feeding with a conveyor belt was implemented to singularize and evenly distribute the seeds. Finally, these subsystems were integrated into a functional prototype. An orthogonal rotation regression experiment was performed with the transport speed, suction negative pressure, and suction height as the experimental factors. The performance was evaluated as the qualified seeding rate, miss-seeding rate, seeding accuracy, and operational efficiency. The experiment revealed that the primary influencing factors on the qualified seeding rate were ranked in the descending order of the suction negative pressure, suction height, and transport speed. The optimal combination of the parameters was identified as a transport speed of 1 000 mm/s, a suction negative pressure of 60 kPa, and a suction height of -1 mm. The better performance was achieved in a high qualified seeding rate of 96.67%, a low miss-seeding rate of 3.33%, and an operational efficiency of 3123 seeds per hour under these optimal conditions. Most importantly, the seeding accuracy was 92.24%. Machine vision was integrated with the high precision of the mechanical orientation. A directional seeding robot was developed for the cucurbit rootstocks. The "bud recognition, posture adjustment, and precise delivery" were combined into a unified system. Directional seeding accuracy and operational efficiency were significantly enhanced to develop the oriented seeding robots in the advanced horticulture and grafting applications. At the same time, the directional seeding equipment can enhance the mechanical grafting efficiency and seeding accuracy.

Planting area of the high value-added crops has been ever increasing in recent years, particularly with the agricultural intensification and scale. However, there is the large accumulation of the insect eggs and pathogenic bacteria in the soil during production, leading to the continuous cropping obstacles, as well as the soil borne pests and diseases. Alternatively, the soil disinfection is one of the most important measures to prevent and control the soil borne diseases and pests. However, the traditional chemical disinfection has often posed the ecological pollution and risks on the human and animal health. Therefore, it is an urgent need to develop the green and pollution-free physical prevention and control technologies. Among them, microwave soil disinfection can be one of the green physical application in sustainable agriculture. The microwave radiation can effectively kill the pests and pathogens in soil, while avoid the residual pollution of chemical agents. Microwave soil disinfection can be expected to fully meet the requirements of green agriculture. This study aims to systematically review the research progress of the microwave soil disinfection. Firstly, the thermal and non-thermal effects of the microwave radiation were introduced for the microwave technology. An analysis was made on its penetration in soil at the different frequency bands (915 and 2 450 MHz) and the application in various depths of soil. Among them, 915 MHz microwave was used to penetrate the soil at a deeper level, while 2 450 MHz microwave was suitable for the shallow disinfection. In addition, the mechanism of the microwave soil interaction was summarized to explore the influence of the microwave working parameters (frequency, power, and exposure time) and soil physical properties on the electromagnetic field distribution and disinfection. And there was the influence of the microwave treatment on the soil physicochemical properties and microbial communities. Furthermore, a comparison was provided on the technological differences between microwave soil disinfection equipment. And among them, the microwave disinfection equipment from developed countries was achieved in the commercial production. While the microwave disinfection equipment in China was still in the experimental prototype stage, indicating the significant gap in development level. Then, some main technological challenges were also proposed in the core components, such as the high-power microwave generators (microwave power supplies, and magnetrons) and intelligent control systems. Microwave disinfection shared some problems at present, including the uncontrolled depth of microwave disinfection, lack of uniformity, absence of core equipment components and intelligent systems, and insufficient ecological safety assessment. Future research should focus on the following directions: 1) Revealing the mechanism of the microwave soil interaction on the non-thermal effects; 2) Breakthrough the key technology of the microwave deep, uniform and efficient disinfection, in order to optimize the energy efficiency; 3) Accelerating the core components, such as the high-power microwave generators and the intelligent upgrade of equipment. This finding can provide the theoretical reference and technical route guidance to optimize microwave soil disinfection.

Plant pigment content is a crucial indicator for assessing photosynthetic efficiency, nutritional status, and physiological health. Its spatial distribution is significantly influenced by variety, location, and environmental factors. However, existing methods for measuring pigment content are often destructive, inefficient, and costly, making them unsuitable for the demands of modern precision agriculture. This study proposes a cross-scale, non-destructive detection method for lettuce pigments by integrating hyperspectral imaging (HSI) technology with deep learning algorithms, addressing the limitations of existing techniques in high-throughput and spatial resolution analysis. In this study, we built a multidimensional dataset based on eight different types of lettuce and developed a deep learning model named LPCNet to predict the contents of chlorophyll a (Chl a), chlorophyll b (Chl b), carotenoids (Car), and total pigment content (TPC) in lettuce. The LPCNet model integrates convolutional neural networks (CNN), bidirectional long short-term memory networks (BiLSTM), and multi-head self-attention (MHSA) mechanisms, enabling automatic extraction of pigment-related key features and simplifying the complex preprocessing and feature selection procedures required in traditional machine learning. Compared to multivariate analysis methods in machine learning, LPCNet demonstrated superior predictive accuracy, with coefficients of determination (R_P²) of 0.9449, 0.8613, 0.9121, and 0.8476 for Chl a, Chl b, Car, and TPC, respectively. Additionally, by combining the hyperspectral reflectance of lettuce canopies with the leaf-level inversion model, we visualized the spatial distribution of pigment content on the canopy of lettuce, achieving cross-scale analysis from leaf to canopy. This study provides an innovative approach for the rapid and accurate assessment of lettuce pigment content and offers an effective visualization tool for revealing the physiological processes and growth development of lettuce.

Maize (Zea mays L.) is a critical staple crop globally, integral to human consumption, food security, and agricultural product stability. The quality and purity of maize seeds, essential for hybrid seed production, are contingent upon effective detasseling. This study investigates the evolution of detasseling technologies and their application in Chinese maize hybrid seed production, with a comparative analysis against the United States. A comprehensive examination of the development and utilization of detasseling technology in Chinese maize hybrid seed production was undertaken, with a specific focus on key milestones. Data from the United States were included for comparative purposes. The analysis encompassed various detasseling methods, including manual, semi-mechanized, and cytoplasmic male sterility, as well as more recent innovations such as detasseling machines, and the emerging field of intelligent detasseling driven by unmanned aerial vehicles (UAVs), computer vision, and mechanical arms. Mechanized detasseling methods were predominantly employed by America. Despite the challenges of inflexible and occasionally overlooked, applying detasseling machines is efficient and reliable. At present, China’s detasseling operations in hybrid maize seed production are mainly carried out by manual work, which is labor-intensive and inefficient. In order to address this issue, China is dedicated to developing intelligent detasseling technology. This study emphasizes the critical role of detasseling in hybrid maize seed production. The United States has embraced mechanized detasseling. The application and development of manual and mechanized detasseling were applied later than those in the United States, but latest intelligent detasseling technologies first appeared in China. Intelligent detasseling is expected to be the future direction, ensuring the quality and efficiency of hybrid maize seed production, with implications for global food security.