Research on Volume Measurement of Logs Based on Embedded Application

Spurred by the worldwide concern for forest protection and increased log sales, most countries have standardized volume calculations to avoid excessive timber protect buyers. However, is currently manually measured, suffering from high labor costs, low measurement progress, imposing significant errors. Thus, automatically obtaining volumetric data of logs a convenient quick solution. Therefore, this work proposes Mask Region Convolutional Neural Network-based (R-CNN) algorithm Logs measurement, named Wood (WM) R-CNN. Specifically, we employ Res2Net structure as backbone obtain receptive fields that exceed input feature size, thus improving our model’s multi-scale information fusion ability. Additionally, WM R-CNN relies on Path Aggregation Feature Pyramid Network’s (PAFPN’s) path enhancement structure, shortening low-level map’s propagation wood contour segmentation accuracy. Extensive experiments Vehicle-mounted Dense (VMDL) dataset demonstrate affords highly appealing accuracy small, medium, large wood, corresponding mAP indicators against current methods 2.0%, 1.2%, 4.4%, respectively. Furthermore, quantitative method based TensorRT compresses proposed model deploy mobile embedded devices. compensate quantization loss, introduce expansion convolution operation manipulate mask map control calculation error all vehicle within 1%. The reveal offers an performance, verifying algorithm’s effectiveness implementation ability terminals.