相对特征在昆虫目级阶元图像自动鉴定中的应用

应用计算机图像技术并结合昆虫分类学原理,设计和开发了一套"昆虫目级阶元标本图像自动鉴定系统",提出了一组基于昆虫标本图像的相对特征,并通过前馈人工神经网络方法进行鉴定测试.测试结果表明,相对特征应用于昆虫目级阶元图像的自动鉴定是有效的,在对7个常见昆虫目的测试中,系统表现出良好的准确性和稳定性,最高正确识别率可以达到95%.     

翅脉的数学形态特征在蝴蝶分类鉴定中的应用研究

能否量化地利用翅脉特征对鳞翅目昆虫进行种类鉴定,这是近年发展起来的数字化昆虫分类鉴定研究中具有创新性的课题。本文使用化学方法去除蝴蝶翅面的鳞片和色斑,通过扫描获取到蝴蝶翅脉图片,利用DrawWing软件对7种蝴蝶前翅内部翅脉交点坐标进行了自动获取。通过计算相邻两点间的距离,利用单变量方差分析和典则判别分析,证明粉蝶科4个属的蝴蝶每个参数对结果都有影响,而绢蝶属3种蝴蝶翅脉交点2~3、7~8和9~1间的距离对其本身的判别无影响。通过分层聚类分析,绢蝶属的3种蝴蝶被聚为一类,在传统昆虫分类学中,它们的关系也最近。此外,通过与半自动提取软件TPSDig提取的数据进行比较,从另一个侧面证明了通过比较翅脉交点间的距离对蝴蝶进行分类鉴定研究具有参考意义。     

几何形态计量学在昆虫自动鉴定中的应用与展望

几何形态计量学着重研究的是生物形态的拓扑结构信息,不受标本大小和形状变化的影响,是进行昆虫自动鉴定的一个新的手段。本文首先介绍了传统形态计量学和几何形态计量学两种昆虫自动鉴定的方法,并着重介绍了轮廓分析法和标志点分析法等几何形态计量学的分析方法,以及目前开发的几何形态计量学昆虫自动鉴定软件,简述了该方法应用于昆虫自动鉴定的工作流程,最后对几何形态计量学在昆虫自动鉴定方面的未来发展进行了展望。     

基于颜色名和OpponentSIFT特征的鳞翅目昆虫图像识别

【目的】本研究旨在探索使用先进的计算机视觉技术实现对昆虫图像的自动分类方法。【方法】通过预处理对采集的昆虫标本图像去除背景,获得昆虫图像的前景蒙板,并由蒙板确定的轮廓计算出前景图像的最小包围盒,剪切出由最小包围盒确定的前景有效区域,然后对剪切得到的图像进行特征提取。首先提取颜色名特征,把原来的RGB（Red-Green-Blue）图像的像素值映射到11种颜色名空间,其值表示RGB值属于该颜色名的概率,每个颜色名平面划分成3×3像素大小的网格,用每格的概率均值作为网格中心点的描述子,最后用空阈金字塔直方图统计的方式形成颜色名视觉词袋特征;其次提取OpponentSIFT(Opponent Scale Invariant Feature Transform)特征,首先把RGB图像变换到对立色空间,对该空间每通道提取SIFT特征,最后用空域池化和直方图统计方法形成OpponentSIFT视觉词袋。将两种词袋特征串接后得到该昆虫图像的特征向量。使用昆虫图像样本训练集提取到的特征向量训练SVM（Support Vector Machine）分类器,使用这些训练得到的分类器即可实现对鳞翅目昆虫的分类识别。【结果】该方法在包含10种576个样本的昆虫图像数据库中进行了测试,取得了100%的识别正确率。【结论】试验结果证明基于颜色名和OpponentSIFT特征可以有效实现对鳞翅目昆虫图像的识别。     

厦门市森林蝴蝶种类

在厦门市森林昆虫调查基础上,1998-2004年对厦门市森林蝴蝶进行种类调查,鉴定厦门市森林蝴蝶7科39种,并编制分种检索表.     

基于F3Net显著性目标检测的蝴蝶图像前背景自动分割

【目的】具有复杂背景的蝴蝶图像前背景分割难度大。本研究旨在探索基于深度学习显著性目标检测的蝴蝶图像自动分割方法。【方法】应用DUTS-TR数据集训练F³Net显著性目标检测算法构建前背景预测模型,然后将模型用于具有复杂背景的蝴蝶图像数据集实现蝴蝶前背景自动分割。在此基础上,采用迁移学习方法,保持ResNet骨架不变,利用蝴蝶图像及其前景蒙板数据,使用交叉特征模块、级联反馈解码器和像素感知损失方法重新训练优化模型参数,得到更优的自动分割模型。同时,将其他5种基于深度学习显著性检测算法也用于自动分割,并比较了这些算法和F³Net算法的性能。【结果】所有算法均获得了很好的蝴蝶图像前背景分割效果,其中,F³Net是更优的算法,其7个指标S测度、E测度、F测度、平均绝对误差(MAE)、精度、召回率和平均IoU值分别为0.940, 0.945, 0.938, 0.024, 0.929,0.978和0.909。迁移学习则进一步提升了F³Net的上述指标值,分别为0.961, 0.964, 0.963, 0.013, 0.965, 0.967和0.938。【结论】研究结果证明结合迁移学习的F³Net算法是其中最优的分割方法。本研究提出的方法可用于野外调查中拍摄的昆虫图像的自动分割,并拓展了显著性目标检测方法的应用范围。     

昆虫自动鉴定技术研究与展望

昆虫自动鉴定是解决目前昆虫鉴定需求不断增多与专业鉴定人员数量相对较少之间矛盾的有效方法之一.本文从昆虫数字图像获取、图像处理、数学特征提取和分类器的设计等方面介绍了昆虫自动鉴定技术的研究进展;并对该项技术的未来发展进行了展望.     

问：蝴蝶翅的颜色是哪里来的？

问：蝴蝶翅的颜色是哪里来的？答：蝴蝶属于昆虫纲鳞翅目,故名思义,蝴蝶的翅是鳞翅。所谓“鳞翅”即膜质剧上覆盖有很多鳞片。此外,在蝴蝶的头部、胸部、腹部和翅基部等处还丛生有介于鳞片和毛之间的鳞毛。蝴蝶的鳞片是由一个单独特化的真皮细胞延伸并穿过表皮扩展而成...     

广西扶绥捕虫记

这次社会实践我们来到了广西扶绥,考察这里最具代表性的物种——白头叶猴,并且了解了那里的地质地貌、植被特征,动物种类等。在这次旅行中,还有一件有意思的活动要数捉昆虫了。为了对昆虫有更多的了解,我们在去捉昆虫之前听了一个讲座。讲座主要讲了蛾子和蝴蝶的区别,老师还拿一些图片让我们通过它们的触角来区分蝴蝶和蛾子。听完了讲座,我对这些平时视而不见的昆虫充满了兴趣,恨不得赶紧到山里去捉昆虫。     

昆虫图像自动鉴别技术

昆虫是地球上物种多样性最为丰富的生物类群,其物种鉴定任务复杂而艰巨,可靠的物种鉴定是开展昆虫学工作的重要基础之一。当前,国内外的人工昆虫物种鉴定能力均不能满足实际需求,因而人们开始不断探索利用计算机自动鉴定昆虫的原理和方法。目前,模式识别技术的迅猛发展已为昆虫图像的自动鉴定提供可能。文章概述昆虫图像自动鉴定技术研究的历史与现状,总结主要原理和方法,介绍工作流程,并展望发展前景。     

Identification of Indian butterflies using Deep Convolutional Neural Network

The conventional butterfly identification method is based on their different morphological characters namely wing-venation, color, shape, patterns and through the dissection studies and molecular techniques which are tedious, expensive and highly time-consuming. To overcome the above aforesaid challenges, a new butterfly identification system using butterfly images has been designed to instantly identify the butterfly with high accuracy. In this study, we construct a new butterfly dataset with 34,024 butterfly images belonging to 315 species from India. We propose and prove the effectiveness of new data augmentation techniques on our dataset. To identify butterflies using photographic images, we built eleven new Deep Convolutional Neural Network (DCNN) butterfly classifier models using eleven pre-trained architectures namely ResNet-18, ResNet-34, ResNet-50, ResNet-121, ResNet-152, Alex-Net, DenseNet-121, DenseNet-161, VGG-16, VGG-19 and SqueezeNet-v1.1. The different model's classification results were compared and the proposed technique achieved a maximum top-1 accuracy(94.44%), top-3 accuracy(98.46%) and top-5 accuracy(99.09%) using ResNet-152 model, followed by DenseNet-161 model achieved the top-1 accuracy(94.31%), top-3 accuracy (98.07%) and top-5 accuracy (98.66%). The results suggest that models can be assertively used to identify butterflies in India.     

昆虫图像分割方法及其应用

昆虫图像自动鉴定是一种快速鉴定昆虫的方法,图像分割则是其中关键步骤。通过搜集和整理国内外近年来针对昆虫图像的分割方法和研究,发现对昆虫图像分割的研究日趋增多。随着计算机图像技术的发展,昆虫图像分割方法吸收了许多图像分割领域中新兴的方法, 诸如采用水平集、边缘流以及结合形状、纹理、色彩等多种要素的智能分割(如JSEG方法)等。虽然大量的图像分割方法被引入到昆虫图像研究中,但是目前分割技术依然是阻碍昆虫图像广泛应用的关键。本文经过总结和分析,发现目前昆虫图像分割研究的往往在各自的测试集上有良好表现, 但是缺乏统一的评价标准, 因此很多方法在昆虫图像中应用难以推广。针对研究中的存在的这些问题,需建立良好的昆虫图像分割评价体系,本文建议通过建立统一的昆虫图像库以及对昆虫图像分割的评价方法深入研究,并且这些工作是当前昆虫图像分割研究亟待完善任务。     

Graph ranking based butterfly segmentation in ecological images

An accurate cultural insect detection and recognition relies mainly on a proper automatic segmentation. This paper deals with butterfly segmentation in ecological images characterized by several artifacts like the complexity of environmental decors and cluttered backgrounds. The distractors contained in the rich ecological environment and the huge difference between butterfly species complicate severely the segmentation and make it a challenging task. As butterflies appears to be well contrasted from their surrounding, we suggest to explore the saliency property to delineate accurately the butterfly boundaries. In this vein, we perform a graph ranking process with high level guidance according to foreground and background cues to improve the quality of segmentation. The ranking accuracy is improved through a weighting scheme that combines accurately color, texture and spatial information. The contribution of each used feature is controlled according to its relevance in highlighting butterfly regions. After that, we initialize foreground seeds from most salient pixels and background seeds from less salient pixels as an input for a Graph-cut algorithm to extract the butterfly from the background. Comparative evaluation has shown that our segmentation scheme outperforms some existing segmentation methods that provide high segmentation scores.     

Taxonomy of Mechanitis (F.) (Lepidoptera: Nymphalidae) from the West Colombian Andes: an Integrative Approach

Species identification in the butterfly genus Mechanitis (F.) (Lepidoptera: Nymphalidae) becomes difficult when it is based only on wing color patterns, a common practice in butterfly taxonomy. Difficulties in Mechanitis taxonomy are related to the widespread mimicry and polymorphism among species belonging to this genus. Species recognition and inventories of Mechanitis genus in geographic areas as the Andean region of Colombia are of particular interest and the use of more than one character for taxonomic identification is desirable. In this study, we included morphological, ecological, and mitochondrial DNA data to identify the occurring species in this region. Species of Mechanitis were studied from ecological, morphological, and molecular perspectives considering host plant identification, oviposition behavior, and life cycles under laboratory conditions. Immature morphology, patterns of wing color, and genital structures of adults were also studied. The genetic barcoding region of the cytochrome oxidase I mitochondrial gene was sequenced and used to verify the limits between species previously defined by the other characters and to validate its usefulness for species delimitation in this particular genus. The integrative approach combining independent datasets successfully allowed species identification as compared to the approach based on a single dataset. Three well-differentiated species were found in the studied region, Mechanitis menapis (Hewitson), Mechanitis polymnia (Linnaeus), and Mechanitis lysimnia (Fabricius). New valuable characters that could improve taxonomic identification in this genus are considered.     

Geometric morphometrics: Current and future in China

Although there were many ancient Chinese mathematicians contributed a lot on geometry,Geometric morphometrics (GM) in modern concept was not firstly proposed by Chinese.The super capability of geometric morphometrics in scientific computing and problem solving has gained a lot of attentions in the world.Until early of 21 centuries,geometric morphometrics was introduced into China.Since then,GM was rapidly applied in many research fields.However,it is a pity that GM is still not well-known in China as many works are published out of China.Thus,the special issue "Geometric morphometrics:Current shape and future directions" is organized.The present issue presents a series of contributions in this scientific field.In future,there will be many considerable new developing fields on GM needed to pay more attentions,for instances,3D geometric morphometrics,4D analysis,visualization of amber,new machine developing,new software developing,automatic identification system,etc.Once these technical bottle-necks on 3D data collecting and merging geometric morphometric data from multiple characters could be solved,the automatic identification system and other fields based on Big Data would come true.     

Artificial intelligence in pest insect monitoring

Abstract Global problems of hunger and malnutrition induced us to introduce a new tool for semi‐automated pest insect identification and monitoring: an artificial neural network system. Multilayer perceptrons, an artificial intelligence method, seem to be efficient for this purpose. We evaluated 101 European economically important thrips (Thysanoptera) species: extrapolation of the verification test data indicated 95% reliability at least for some taxa analysed. Mainly quantitative morphometric characters, such as head, clavus, wing, ovipositor length and width, formed the input variable computation set in a Trajan neural network simulator. The technique may be combined with digital image analysis.     

A method for reconstructing three-dimensional positions of swarming mosquitoes

We developed a new field method for reconstructing the three-dimensional positions of swarming mosquitoes. This method overcame certain inherent difficulties accompanied by conventional stereoscopic methods and is applicable to three-dimensional measurements of other insect species. Firstly, we constructed a probabilistic model for stereoscopy; if mosquitoes and six reference points with known coordinates were photographed simultaneously from two or more perspectives, then from the positions of images of mosquitoes and the reference points on the photographs, 1) the position of each camera with respect to the reference points is estimated; 2) stereo images which correspond to an identical real mosquito are matched; and 3) the spatial positions of the mosquitoes are determined. We automated the processes 1), 2) and 3), developing computer programs based on our model. We then constructed a portable system for three-dimensional measurements of swarming mosquitoes in the field. Initial data that illustrate the application of our method to studying mosquito swarming were presented.