Search

The key technology of automated apple harvesting is detecting apples quickly and accurately. The traditional detection methods of apple detection are often slow and inaccurate in unstructured orchards. Therefore, this article proposes an improved YOLOv5s-GBR model for orchard apple detection under complex natural conditions. First, the researchers collected photos of apples in their natural environments from different angles; then, we enhanced the dataset by changing the brightness, rotating the images, and adding noise. In the YOLOv5s network, the following modules were introduced to improve its performance: First, the YOLOv5s model’s backbone network was swapped out for the GhostNetV2 module. The goal of this improvement was to lessen the computational burden on the YOLOv5s algorithm while increasing the detection speed. Second, the bi-level routing spatial attention module (BRSAM), which combines spatial attention (SA) with bi-level routing attention (BRA), was used in this study. By strengthening the model’s capacity to extract important characteristics from the target, its generality and robustness were enhanced. Lastly, this research replaced the original bounding box loss function with a repulsion loss function to detect overlapping targets. This model performs better in detection, especially in situations involving occluded and overlapping targets. According to the test results, the YOLOv5s-GBR model improved the average precision by 4.1% and recall by 4.0% compared to those of the original YOLOv5s model, with an impressive detection accuracy of 98.20% at a frame rate of only 101.2 fps. The improved algorithm increases the recognition accuracy by 12.7%, 10.6%, 5.9%, 2.7%, 1.9%, 0.8%, 2.6%, and 5.3% compared to those of YOLOv5-lite-s, YOLOv5-lite-e, yolov4-tiny, YOLOv5m, YOLOv5l, YOLOv8s, Faster R-CNN, and SSD, respectively, and the YOLOv5s-GBR model can be used to accurately recognize overlapping or occluded apples, which can be subsequently deployed in picked robots to meet the realistic demand of real-time apple detection.

In recent years, two-dimensional (2D) type-II van der Waals (vdW) heterojunctions have emerged as promising candidates for high-performance photodetectors. However, direct experimental evidence confirming the enhancement of photoelectric properties by the heterojunction’s type and structure remains scarce. In this work, we present flexible photodetectors based on individual GaSe and MoS2, as well as a vertically stacked type-II GaSe/MoS2 vdW heterojunction on polyethylene terephthalate (PET) substrate. These devices demonstrate outstanding responsivities and rapid response speeds, ensuring stable and repeatable light detection. It is notable that the GaSe/MoS2 heterojunction photodetector exhibits the highest on-off ratio and fastest response speed, attributed to the formation of type-II band alignment. Furthermore, the GaSe/MoS2 heterojunction photodetector maintains robust stability even in a bent state, highlighting remarkable flexibility. This work exemplifies the type-II vdW heterojunctions in enhancing photoelectric properties through direct in-situ experimentation, laying the groundwork for practical applications of 2D flexible photodetectors.

In recent years, two-dimensional (2D) layered transition metal dichalcogenides (TMDs) have received significant attention due to their exceptional electrical and optical properties. Among these 2D materials, ReS2 distinguishes itself through its unique optical and conductance anisotropy. Despite concerted efforts to produce high-quality ReS2, the unique interlayer decoupling properties pose substantial challenges in growing large-area ReS2 thin films, with the preparation of single layers proving even more complex. In this work, large-scale continuous monolayer and bilayer ReS2 films were successfully grown on mica substrates using low-pressure chemical vapor deposition (LPCVD). Photodetectors were fabricated using the prepared high-quality ReS2 films, and the devices presented stable photoresponse and enhanced response sensitivity. The production of continuous ReS2 atomic layers heralds promising prospects for large-scale integrated circuits and advances the practical application of optoelectronics based on 2D layered materials.

Monolayer MoS2 has emerged as a highly promising candidate for next-generation electronics. However, the production of monolayer MoS2 with a high yield and low cost remains a challenge that impedes its practical application. Here, a significant breakthrough in the batch production of wafer-scale monolayer MoS2 via chemical vapor deposition is reported. Notably, a single preparation process enables the growth of multiple wafers simultaneously. The homogeneity and cleanliness of the entire wafer, as well as the consistency of different wafers within a batch, are demonstrated via morphology characterizations and spectroscopic measurements. Field-effect transistors fabricated using the grown MoS2 exhibit excellent electrical performances, confirming the high quality of the films obtained via this novel batch production method. Additionally, we successfully demonstrate the batch production of wafer-scale oxygen-doped MoS2 films via in situ oxygen doping. This work establishes a pathway towards mass preparation of two-dimensional materials and accelerates their development for diverse applications.