《電子技術(shù)應(yīng)用》
您所在的位置:首頁 > 通信与网络 > 设计应用 > 基于双线性池化的实蝇分类注意力网络
基于双线性池化的实蝇分类注意力网络
2023年电子技术应用第5期
彭莹琼1,2,俞融融3,尹乘乐4,洪恩松2,俞小明3,赵雷3,何雯洁2,邓泓1,2
(1.江西农业大学 江西省高等学校农业信息技术重点实验室与软件研究所,江西 南昌330045; 2.江西农业大学 软件学院,江西 南昌330045;3.江西农业大学 计算机与信息工程学院,江西 南昌330045; 4.德布勒森大学,匈牙利 德布勒森4032)
摘要: 实蝇是国内外备受关注的检疫害虫,种类繁多。不同种类的实蝇外形大小相似,不易鉴别。此外,在实际应用中,鉴别实蝇的可用信息会受遮挡、视角、光影变幻等因素影响,导致实蝇自动识别工作难以进行。提出基于双线性池化的实蝇分类注意力网络,用于学习有效的实蝇鉴别特征。该网络由显著性特征模块和跨层双线性模块两个部分组成:显著性特征模块通过对不同卷积层进行滤波增强处理,实现特征增强;跨层双线性模块基于双线性池化融合特征,确定注意部位,挖掘判别特征。在具有自然环境背景的实蝇数据集上进行的实验表明,该方法效果较好,具有良好的实际应用前景。
中圖分類號:TP391.41;TP18
文獻(xiàn)標(biāo)志碼:A
DOI: 10.16157/j.issn.0258-7998.233817
中文引用格式: 彭瑩瓊,俞融融,尹乘樂,等. 基于雙線性池化的實(shí)蠅分類注意力網(wǎng)絡(luò)[J]. 電子技術(shù)應(yīng)用,2023,49(5):8-13.
英文引用格式: Peng Yingqiong,Yu Rongrong,Ying Chengle,et al. Attention networks for fruit fly classification based on bilinear pooling[J]. Application of Electronic Technique,2023,49(5):8-13.
Attention networks for fruit fly classification based on bilinear pooling
Peng Yingqiong1,2,Yu Rongrong3,Ying Chengle4,Hong Ensong2,Yu Xiaoming3,Zhao Lei3,He Wenjie2,Deng Hong1,2
(1.The Colleges and Universities of Jiangxi Province for Key Laboratory of Information Technology in Agriculture and Software Institute, Jiangxi Agricultural University, Nanchang 330045, China; 2.College of Software, Jiangxi Agricultural University, Nanchang 330045, China; 3.College of Computer and Information Engineering, Jiangxi Agricultural University, Nanchang 330045, China; 4.University of Debrecen,Debrecen 4032,Hungary)
Abstract: Fruit fly is a kind of quarantine pest that attracts much attention at home and abroad. There are many kinds of fruit flies. Different kinds of fruit flies are similar in shape and size, which is difficult to identify. In addition, in practical applications, it is difficult to identify fruit flies due to the lack of information about shielding, view-point, changing light and shadow and other factors. This study proposes a bilinear pooled attention network for fruit fly classification to learn effective discriminant characteristics. The network is composed of two parts: saliency feature module and cross-layer bilinear feature module. Saliency feature module realizes feature enhancement by filtering enhancement processing of two different convolution layers. Cross-layer bilinear module is based on bilinear pooling fusion features, determines the attention location, and mines discriminant features. Experiments on fruit fly’s data set with natural environment background show that the method is effective and has good practical application prospect.
Key words : fruit fly detection;bilinear pooling;attention mechanism

0 引言

實(shí)蠅作為亞太地區(qū)重要的檢疫性害蟲,具有寄主多、蟲害擴(kuò)散迅速的特點(diǎn)。該類害蟲能夠寄生于橘、桃、番石榴、楊梅等46個科 250多種水果、蔬菜和花卉。在不注重防控的情況下,實(shí)蠅能輕易造成80%到100%的損失。以福建省為例,該省2016年受瓜實(shí)蠅、具條實(shí)蠅等實(shí)蠅害蟲影響,導(dǎo)致約313.48億元經(jīng)濟(jì)損失,其中直接經(jīng)濟(jì)損失約221.74億元,包括生態(tài)損失在內(nèi)的間接經(jīng)濟(jì)損失約為7.5億元。所以防治實(shí)蠅對于減少農(nóng)業(yè)經(jīng)濟(jì)損失起重要作用。針對實(shí)蠅的檢疫與防治,此研究將準(zhǔn)確識別實(shí)蠅個體類別作為首要任務(wù),為實(shí)蠅檢測的實(shí)際應(yīng)用提供思路。


本文詳細(xì)內(nèi)容請下載:http://m.ihrv.cn/resource/share/2000005300




作者信息:

彭瑩瓊1,2,俞融融3,尹乘樂4,洪恩松2,俞小明3,趙雷3,何雯潔2,鄧泓1,2

(1.江西農(nóng)業(yè)大學(xué) 江西省高等學(xué)校農(nóng)業(yè)信息技術(shù)重點(diǎn)實(shí)驗(yàn)室與軟件研究所,江西 南昌330045;2.江西農(nóng)業(yè)大學(xué) 軟件學(xué)院,江西 南昌330045;3.江西農(nóng)業(yè)大學(xué) 計算機(jī)與信息工程學(xué)院,江西 南昌330045;4.德布勒森大學(xué),匈牙利 德布勒森4032)


微信圖片_20210517164139.jpg

此內(nèi)容為AET網(wǎng)站原創(chuàng),未經(jīng)授權(quán)禁止轉(zhuǎn)載。

相關(guān)內(nèi)容