In orchards, there are many tasks that have not yet been automated, such as harvesting, transportation, weeding, and monitoring. In this study, we attempted 3D mapping and automatic driving in a chestnut orchard using simultaneous localization and mapping (SLAM) with 3D-LiDAR and an IMU. As a result, parameters suitable for normal distribution transform (NDT) matching in a chestnut orchard were identified. In automatic driving, the lateral error of the path was less than 10 cm, which was sufficient for an agricultural robot to travel in an orchard. The effect of the abundance of foliage on the accuracy of self-position estimation was small, indicating the reusability of the created 3D map.

View full abstract

A bibliometric analysis of articles published between 1963 and 2022 elucidated the trajectory of heavy metal contamination as a subject in food safety research. China emerged as the leading contributor, with significant publications addressing heavy metals’ multifaceted impact on food safety. Top author keywords over six decades underscored the concern about specific metals, particularly cadmium, and their accumulation in food sources. The prominence of terms like ‘biomaterials’ and ‘biosensor’ showcased the shift towards innovative detection and mitigation techniques. Three primary clusters of research were identified: soil contamination, heavy metals in various foods and advanced detection techniques. This analysis offers vital insights, identifying prevalent themes and prospective directions for future research efforts.

View full abstract

Computer vision and deep learning are one of the main technologies for weed intelligent recognition in farmland. However, when using the deep learning technology to identify weeds in the field broomcorn millet growing at seedling stage, the weeds grow densely, which not only leads to the imbalance of positive and negative sample data, but also causes the problem of small area segmentation. In this paper, the combination of dice and focal loss was applied to Segformer semantic segmentation network with MiT-B3 as encoder block to identify intensive weeds in seedling broomcorn millet fields. The accuracy scores on the testing set were 95.23 %. The results show that the method proposed in this paper can effectively identify the seedling intensive weeds.

View full abstract

Extreme environments, like the steep olive groves in Liguria, Italy, cannot be reached by tractors and large-sized devices. This paper describes a small, tracked elevation platform able to lift the farmers close to the branches for harvesting or pruning. The vehicle moves thanks to tracks. The elevation platform, having no motors and no sensors, is powered by hand. A prototype of the forest lift has been tested on the field. The forest lift has a maximum elevation of 2 m with a tilting (0 ° + 30 °) and rolling mechanism (-15 ° + 15 °) compensating for steep terrains. The iron prototype weighs 400 kg and is 3,160 mm tall, 2,000 mm long and 900 mm wide.

View full abstract