基于编解码网络的生猪骨架提取方法研究  

RESEARCH OF PIG SKELETON EXTRACTION METHOD BASED ON ENCODER-DECODER NETWORK

作　　者：王泽华 徐爱俊[2] 周素茵[2] 叶俊华 夏芳 Wang Zehua;Xu Aijun;Zhou Suyin;Ye Junhua;Xia Fang(College of Mathematics and Computer Science,Zhejiang Agriculture and Forestry University,Hangzhou 311300,Zhejiang,China;Key Laboratory of Forestry Intelligent Monitoring and Information Technology of Zhejiang Province,Hangzhou 311300,Zhejiang,China;School of Environmental Science and Technology,Zhejiang Agriculture and Forestry University,Hangzhou 311300,Zhejiang,China;Digital Country Institute,Zhejiang Agriculture and Forestry University,Hangzhou 311300,Zhejiang,China)

机构地区：[1]浙江农林大学数学与计算机学院,浙江杭州311300 [2]浙江农林大学林业感知技术与智能装备国家林业与草原局重点实验室,浙江杭州311300 [3]浙江农林大学环境与资源学院,浙江杭州311300 [4]浙江农林大学数字乡村研究所,浙江杭州311300

出　　处：《计算机应用与软件》2025年第4期181-188,共8页Computer Applications and Software

基　　金：浙江省农业重大技术协同推广计划项目(2021XTTGXM01-02);浙江省公益技术应用研究计划项目(LGN19F010001)。

摘　　要：针对生猪骨架提取难度大、精度低、耗时长等问题,提出一种基于编解码网络的生猪骨架提取方法。该文构建关键点热力图生成模型,将ResNet50残差网络和U-Net语义分割网络相结合,搭建编码-解码网络结构并引入注意力机制,以提高尾、蹄等小目标关键点的特征提取精度;在生成关键点热力图的同时预测关键点偏移量,弥补反算关键点原始位置时的精度损失,再利用霍夫投票机制对二者进行加权聚合,最终映射得到生猪骨架。实验结果表明,骨架提取准确率为85.27%。相较于ResNet50残差网络,在耗时相近的情况下,准确率提高了22.67个百分点。该研究为生猪骨架提取提供了一种新的方法,可为进一步开展生猪行为研究提供技术参考。Aimed at the problems of pig skeleton extraction,such as difficulty,low accuracy and long-time consumption,a pig skeleton extraction method based on encoder-decoder network is proposed.The key point heat map generation model was constructed,ResNet50 residual network and U-Net semantic segmentation network were combined to build an encoder-decoder network structure,and the attention mechanism was introduced to improve the feature extraction accuracy of the key points of small targets such as tail and hoof.The offset of key points was predicted while generating the key point heat map,which made up for the accuracy loss when calculating the original position of the key points.The Hough voting mechanism was used to weighted aggregate the two points,and the pig skeleton was finally mapped.The experimental results show that the skeleton extraction accuracy is 85.27%.Compared with the ResNet50 residual network,the accuracy is increased by 22.67 percentage points with similar time consumption.This study provides a new method for pig skeleton extraction,which can provide a technical reference for further pig behavior research.

关 键 词：骨架提取 关键点检测 生猪 注意力机制 特征提取 编解码网络 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]