**5. Conclusions**

We propose an intelligent forklift cargo accurate transfer system, which consists of three main parts: pallet monitoring module, pallet positioning module, and highprecision control module. The system creatively introduces small target detection into pallet recognition to improve the recognition rate; using Yolov5-based pallet pose detection algorithm to improve the coverage and recognition accuracy of the algorithm. For the whole system, we proved its effectiveness and reliability by actually collecting a large amount of data. Among them, the recognition correct rate of pallet monitoring module reaches more than 99.5% in 7 days continuous experiments, and the insertion error of the intelligent forklift is below ±6 mm after 1000 experiments through pallet positioning and control algorithm. The performance of the whole system is greatly higher than the existing common systems, and has been used in factories and warehousing environments on the ground. We will further investigate the use of lightweight networks to refine our system in order to reduce computational resource consumption and computation time. In this way, the cost of commercializing the system can be further reduced.

**Author Contributions:** Conceptualization, J.R. and Y.P.; methodology, J.R.; software, Y.H.; validation, J.R., P.Y. and Y.H.; formal analysis, Y.P.; writing—original draft preparation, J.R.; writing—review and editing, Y.P. and Z.X.; supervision, W.G. and Z.X.; project administration, Y.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Zhejiang Province Science and Technology Plan Project (2022C01214).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data will be made available upon request from the authors.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

