 This paper presents a novel 3D reconstruction method using multi-angle point clouds from a binocular depth camera and a YOLO-based Nero model to improve the accuracy of contactless grain moisture monitoring in actual use. The study developed an embedded and low-cost monitoring system for in-warehouse grain bags, which predicted targets 3D size and boosted contactless grain moisture measuring. The method fused multi-angle point clouds using a rotation matrix and deployed all the above methods on a raspberry pie-embedded board to perform grain bags 3D reconstruction and size prediction. Results demonstrated that the NCN accelerated model significantly enhanced the average processing speed, nearly 30 times faster than the torch model. The proposed system predicted the object's length, width, and height with accuracies of 97.76%, 97.02%, and 96.81%, respectively. In the future, the system will mount three depth cameras for achieving real-time size prediction and introduce a contactless measuring tool to finalize grain moisture detection. This article was authored by Xu Jinghua, Sumao, Dong Dai, and others. We are article.tv, links in the description below.