基于改进ConvNeXt模型的黄羽鸡表皮层黑色素智能分级方法  

Grading melanin in epidermal layer of yellow feather broiler using improved ConvNeXt model

作　　者：肖德琴[1,2] 陈芳玲 刘又夫 谢青梅[3] 张新珩 XIAO Deqin;CHEN Fangling;LIU Youfu;XIE Qingmei;ZHANG Xinheng(Key Laboratory of Smart Agricultural Technology in Tropical South China Ministry of Agriculture and Rural Affairs,South China Agricultural University,Guangzhou 510642,China;College of Mathematics and Information,South China Agricultural University,Guangzhou,510642,China;College of Animal Science,South China Agricultural University,Guangzhou,510642,China)

机构地区：[1]华南农业大学农业农村部华南热带智慧农业技术重点实验室,广州510642 [2]华南农业大学数学与信息学院,广州510642 [3]华南农业大学动物科学学院,广州510642

出　　处：《农业工程学报》2024年第3期165-174,共10页Transactions of the Chinese Society of Agricultural Engineering

基　　金：江苏省重点研发计划(现代农业)重点项目(BE2022379);国家现代农业产业技术体系(CARS-42-13);岭南现代农业(DT20220003)。

摘　　要：为解决活体黄羽鸡表皮层黑色素分级方法成本高、效率低下、分级环境易受环境光干扰等问题,该研究探索一种基于ConvNeXt模型的黄羽鸡表皮层黑色素智能分级方法ConvNeXt-WPCA,用于实现活体黄羽鸡表皮层黑色素智能分级。ConvNeXt-WPCA模型通过以下3点改进提高模型对黄羽鸡黑色素的识别效果:1)针对黄羽鸡黑色素图像RGB三通道内黑色素信息分布不均衡问题,改变输入图片通道权重来增强模型对黑色素特征的提取能力;2)使用部分卷积代替深度可分离卷积,减少模型计算量和内存访问次数提高对计算资源的利用率;3)引入坐标注意力机制,引导模型关注黄羽鸡胸腹部及肛门附近皮肤提升模型精度。同时,该研究还设计一种双光源图像获取装置,分别在自然光和偏振光条件下拍摄黄羽鸡样本,以减小分级结果受环境光干扰的影响,并探索偏振光在黑色素分级任务中的应用潜力。结果表明ConvNeXt-WPCA模型相较标准ConvNeXt模型,针对自然光下黄羽鸡黑色素图像数据集分级准确率提升9.68个百分点,最终达到89.03%的识别准确率,针对偏振光下黄羽鸡黑色素图像数据集分级准确率提升15.26个百分点,最终达到98.87%的识别准确率。该研究证实基于偏振光条件获取的黄羽鸡表皮层黑色素图像分级效果优于自然光条件,提出的ConvNeXt-WPCA黄羽鸡表皮层黑色素分级方法识别准确率高,同时模型参数量及浮点计算量均有降低,为黄羽鸡表皮层黑色素智能分级实际应用提供了理论基础及技术支持。Melanin grading has been widely used in the epidermal layer of live yellow feather broilers.However,it is in high demand to improve the efficiency,cost-saving and susceptibility to lighting conditions.This study aims to explore an intelligent melanin grading(ConvNeXt-WPCA)for the epidermal layer of live yellow feather broilers using the ConvNeXt model.Three key enhancements were proposed to improve the ConvNeXt-WPCA model for the recognition of melanin in broilers.Firstly,the channel weights of the input images were adjusted to treat the uneven distribution of melanin across the RGB channels in the melanin images of yellow feather broilers.The channel with more melanin was then emphasized to extract the melanin features.The reweighting of channels was used to more effectively aggregate the melanin signals for the better classification performance of the deep learning model.Secondly,Depthwise Separable Convolution(DWConv)was replaced with the partial convolution.The computational load and memory access times were reduced to improve the utilization of computational resources.Lastly,the Coordinate Attention(CA)module was introduced to focus on the key skin regions near the chest and anus of yellow feather broiler,thereby improving the classification accuracy of models.At the same time,a dual-light source image acquisition device was designed to efficiently and simultaneously collect images under both normal and polarized lighting conditions.Sufficient data was available for model training and performance evaluation.There was a minimum impact of lighting conditions on grading.Furthermore,the potential application of polarized light was also explored in the tasks of melanin grading.The results demonstrated that the ConvNeXt-WPCA model was improved by 9.68 percentage points in the grading accuracy rate of the melanin image dataset for the yellow feather broiler under natural light,compared with the standard ConvNeXt model.A final recognition accuracy rate reached 89.03%.The grading accuracy rate was improved by 15.26 p

关 键 词：模型 深度学习 注意力机制 黄羽鸡 ConvNeXt 品质分级 偏振光 

分 类 号：S126[农业科学—农业基础科学]