基于改进YOL011的木材端面识别模型设计

打开文本图片集

中图分类号：S776 文献标识码：A DOI： 10.7525/j.issn.1006-8023.2026.01.007

Wood Log End Recognition Model Design Based on Improved YOLO11

ZHANG Xiaobo， ZENG Zirong*，LIAO Caixia （SchoolfAutomotiehnicandElectronics，angiEnviromentalEngeeingVocatioalCog，Ganu3i）

Abstract：Natural wood end surfaces exhibitirregular textures and defect features，making end surfacerecognition and localization a challenging problem.To enhance detection accuracy while reducing model parameters and improving computational eficiencyfor mobiledeployment，this study proposes animprovedend-to-end deep learning modeltailored for log detectionby enhancing theYOLO11architecture.Firstly，thePP-LCNetbackbone isadopted toreplace theoriginal YOLO11 backbone，efectivelyreducing the number of parameters，expanding the receptive field，and improving large targetdetection precision.Secondly，a parameter-free atention mechanism，SimAM，is integrated intotheneck network toadaptively emphasize criticalfeaturesandsuppressredundant information，therebyenhancingsmalltargetrecognition capabilities.Finally，the nrmalized Waserstein distance（NWD）loss function is introduced，which is more suitable formeasuring similaritybetween extremelysmalltargets，furtherimproves theaccuracyand precisionof woodend surface identification.Experimentalresultsdemonstrate thattheimproved modelachieves higherendsurfacerecognitionaccuracy compared to the baseline model，the improved model improves 2. 65% and 5.29% on the mAP@0.5 and mAP @ （204 0.95 metrics，and FLOPs are decreased by 15. 15% . It has good application value in the field of log volume measurement. Kavworde. Inn timhar·and-curfaoa raononitinn·daan larning. YOI O anhanoamant·nhiant datantinn

0 引言

森林资源采伐与运输环节中，原木楞堆作为临时存储与转运的核心载体，其材积计量长期依赖人工检尺手段。（剩余16540字）