高分辨卫星图像卷积神经网络分类模型

目的 卫星图像往往目标、背景复杂而且带有噪声,因此使用人工选取的特征进行卫星图像的分类就变得十分困难。提出一种新的使用卷积神经网络进行卫星图像分类的方案。使用卷积神经网络可以提取卫星图像的高层特征,进而提高卫星图像分类的识别率。方法 首先,提出一个包含六类图像的新的卫星图像数据集来解决卷积神经网络的有标签训练样本不足的问题。其次,使用了一种直接训练卷积神经网络模型和3种预训练卷积神经网络模型来进行卫星图像分类。直接训练模型直接在文章提出的数据集上进行训练,预训练模型先在ILSVRC（the ImageNet large scale visual recognition challenge）-2012数据集上进行预训练,然后在提出的卫星图像数据集上进行微调训练。完成微调的模型用于卫星图像分类。结果 提出的微调预训练卷积神经网络深层模型具有最高的分类正确率。在提出的数据集上,深层卷积神经网络模型达到了99.50%的识别率。在数据集UC Merced Land Use上,深层卷积神经网络模型达到了96.44%的识别率。结论 本文提出的数据集具有一般性和代表性,使用的深层卷积神经网络模型具有很强的特征提取能力和分类能力,且是一种端到端的分类模型,不需要堆叠其他模型或分类器。在高分辨卫星图像的分类上,本文模型和对比模型相比取得了更有说服力的结果。     

基于迁移学习及气象卫星云图的台风等级分类研究

针对传统卫星云图特征提取方法复杂且深度卷积神经网络（Deep Convolutional Neural Network, DCNN）模型开发困难的问题,提出一种基于参数迁移的台风等级分类方法。利用日本气象厅发布的近40 a 10 000多景台风云图数据,构建了适应于迁移学习的台风云图训练集和测试集。在大规模ImageNet源数据集上训练出3种源模型VGG16,InceptionV3和ResNet50,依据台风云图低层特征与高层语义特征的差异,适配网络最佳迁移层数并冻结低层权重,高层权重采用自适应微调策略,构建出了适用于台风小样本数据集的迁移预报模型T-typCNNs。实验结果表明:T-typCNNs模型在自建台风数据集上的训练精度为95.081%,验证精度可达91.134%,比利用浅层卷积神经网络训练出的精度高18.571%,相比于直接用源模型训练最多提高9.819%。     

Research on Transfer Learning Methods for Classification of Typhoon Cloud Image

Aiming at the complexity of traditional methods for feature extraction about satellite cloud images, and the difficulty of developing deep convolutional neural network from scratch, a parameter-based transfer learning method for classifying typhoon intensity is proposed. Take typhoon satellite cloud images published by Japan Meteorological Agency, which includes 10 000 scenes among nearly 40 years to construct training and test typhoon datasets. Three deep convolutional neural networks, VGG16, InceptionV3 and ResNet50 are trained as source models on the large-scale ImageNet datasets. Considering the discrepancy between low-level features and high-level semantic features of typhoon cloud images, adapt the optimal number of transferable layers in neural networks and freeze weights of low-level network. Meanwhile, fine-tune surplus weights on typhoon dataset adaptively. Finally, a transferred prediction model which is suitable for small sample typhoon datasets, called T-typCNNs is proposed. Experimental results show that the T-typCNNs can achieve training accuracy of 95.081% and testing accuracy of 91.134%, 18.571% higher than using shallow convolutional neural network, 9.819% higher than training with source models from scratch.     

基于深度卷积神经网络的物体识别算法

针对传统物体识别算法中人工设计出来的特征易受物体形态多样性、光照和背景的影响,提出了一种基于深度卷神经网络的物体识别算法。该算法基于NYU Depth V2场景数据库,首先将单通道深度信息转换为三通道;再用训练集中的彩色图片和转换后的三通道深度图片分别微调两个深度卷积神经网络模型;然后用训练好的模型对重采样训练集中的彩色和深度图片提取模型第一个全连接层的特征,并将两种模态的特征串联起来,训练线性支持向量机（LinSVM）;最后将所提算法应用到场景理解任务中的超像素特征提取。所提方法在测试集上的物体分类准确度可达到91.4%,比SAE-RNN方法提高4.1个百分点。实验结果表明所提方法可提取彩色和深度图片高层特征,有效提高物体分类准确度。     

基于多通道分离整合的多尺度单幅图像去雨算法

由于密度不同的雨对图像造成的遮挡不同,图像去雨一直都是一项极具挑战性的任务。目前,基于深度学习的图像去雨算法已经成为主流。然而,多数深度学习的架构都是通过堆叠卷积层来设计的,执行去雨任务后图像仍存在着大小不一的雨痕,这些方法并不能很好地关注训练中雨图的局部信息和上下文信息。为了解决上述问题,本文设计一种基于多通道分离整合的卷积神经网络用于图像去雨。第一步通过通道分离,再利用卷积层间的层级连接,构成多尺度模块,最终将不同通道的输出进行整合。该模块可以增大感受野,探索特征图之间的空间信息,更好地提取特征。第二步利用渐进网络来反复计算挖掘上下文信息,能够很好关联到全局特征。整体模型易于实施,可以端对端训练。在常用的数据集以及自建的自动驾驶雨天数据集上的大量实验表明,本文方法比现有方法取得了明显的改进。     

遥感图像飞机目标分类的卷积神经网络方法

目的 遥感图像飞机目标分类,利用可见光遥感图像对飞机类型进行有效区分,对提供军事作战信息有重要意义。针对该问题,目前存在一些传统机器学习方法,但这些方法需人工提取特征,且难以适应真实遥感图像的复杂背景。近年来,深度卷积神经网络方法兴起,网络能自动学习图像特征且泛化能力强,在计算机视觉各领域应用广泛。但深度卷积神经网络在遥感图像飞机分类问题上应用少见。本文旨在将深度卷积神经网络应用于遥感图像飞机目标分类问题。方法 在缺乏公开数据集的情况下,收集了真实可见光遥感图像中的8种飞机数据,按大致4∶1的比例分为训练集和测试集,并对训练集进行合理扩充。然后针对遥感图像与飞机分类的特殊性,结合深度学习卷积神经网络相关理论,有的放矢地设计了一个5层卷积神经网络。结果 首先,在逐步扩充的训练集上分别训练该卷积神经网络,并分别用同一测试集进行测试,实验表明训练集扩充有利于网络训练,测试准确率从72.4%提升至97.2%。在扩充后训练集上,分别对经典传统机器学习方法、经典卷积神经网络LeNet-5和本文设计的卷积神经网络进行训练,并在同一测试集上测试,实验表明该卷积神经网络的分类准确率高于其他两种方法,最终能在测试集上达到97.2%的准确率,其余两者准确率分别为82.3%、88.7%。结论 在少见使用深度卷积神经网络的遥感图像飞机目标分类问题上,本文设计了一个5层卷积神经网络加以应用。实验结果表明,该网络能适应图像场景,自动学习特征,分类效果良好。     

Hatching eggs classification based on deep learning

In order to realize the fertility detection and classification of hatching eggs, a method based on deep learning is proposed in this paper. The 5-days hatching eggs are divided into fertile eggs, dead eggs and infertile eggs. Firstly, we combine the transfer learning strategy with convolutional neural network (CNN). Then, we use a network of two branches. In the first branch, the dataset is pre-trained with the model trained by AlexNet network on large-scale ImageNet dataset. In the second branch, the dataset is directly trained on a multi-layer network which contains six convolutional layers and four pooling layers. The features of these two branches are combined as input to the following fully connected layer. Finally, a new model is trained on a small-scale dataset by this network and the final accuracy of our method is 99.5%. The experimental results show that the proposed method successfully solves the multi-classification problem in small-scale dataset of hatching eggs and obtains high accuracy. Also, our model has better generalization ability and can be adapted to eggs of diversity.     

Automatic segmentation of optic disc in retinal fundus images using semi-supervised deep learning

Diseases of the eye require manual segmentation and examination of the optic disc by ophthalmologists. Though, image segmentation using deep learning techniques is achieving remarkable results, it leverages on large-scale labeled datasets. But, in the field of medical imaging, it is challenging to acquire large labeled datasets. Hence, this article proposes a novel deep learning model to automatically segment the optic disc in retinal fundus images by using the concepts of semi-supervised learning and transfer learning. Initially, a convolutional autoencoder (CAE) is trained to automatically learn features from a large number of unlabeled fundus images available from the Kaggle’s diabetic retinopathy (DR) dataset. The autoencoder (AE) learns the features from the unlabeled images by reconstructing the input images and becomes a pre-trained network (model). After this, the pre-trained autoencoder network is converted into a segmentation network. Later, using transfer learning, the segmentation network is trained with retinal fundus images along with their corresponding optic disc ground truth images from the DRISHTI GS1 and RIM-ONE datasets. The trained segmentation network is then tested on retinal fundus images from the test set of DRISHTI GS1 and RIM-ONE datasets. The experimental results show that the proposed method performs on par with the state-of-the-art methods achieving a 0.967 and 0.902 dice score coefficient on the test set of the DRISHTI GS1 and RIM-ONE datasets respectively. The proposed method also shows that transfer learning and semi-supervised learning overcomes the barrier imposed by the large labeled dataset. The proposed segmentation model can be used in automatic retinal image processing systems for diagnosing diseases of the eye.

     

基于深度卷积神经网络End-to-End模型的亲属关系认证算法

针对如何利用人脸图像进行亲属关系认证的问题,提出基于深度卷积神经网络End-to-End模型的亲属关系认证算法.首先,构建一个包含卷积层、全连接层和soft-max分类层的深度卷积神经网络模型.卷积层可以提取亲子图像的隐性特征,全连接层可以将提取的隐性特征映射为亲属关系认证的二分类问题,soft-max分类层可以直接判断该对样本是否具有亲属关系.然后,将成对的标记训练数据输入网络进行迭代,优化深度网络模型参数,直至损失曲线稳定.最后,利用训练完毕的深度网络模型对输入测试图像对进行分类判决,通过统计得到最终的准确率.在KinFaceWI和KinFaceWII数据库上的结果显示,相比以往的亲属关系认证算法,文中模型具有更好的性能.     

多级卷积特征金字塔的细粒度食物图片识别

目的 食物图片具有结构多变、背景干扰大、类间差异小、类内差异大等特点,比普通细粒度图片的识别难度更大。目前在食物图片识别领域,食物图片的识别与分类仍存在精度低、泛化性差等问题。为了提高食物图片的识别与分类精度,充分利用食物图片的全局与局部细节信息,本文提出了一个多级卷积特征金字塔的细粒度食物图片识别模型。方法 本文模型从整体到局部逐级提取特征,将干扰较大的背景信息丢弃,仅针对食物目标区域提取特征。模型主要由食物特征提取网络、注意力区域定位网络和特征融合网格3部分组成,并采用3级食物特征提取网络的级联结构来实现特征由全局到局部的转移。此外,针对食物图片尺度变化大的特点,本文模型在每级食物特征提取网络中加入了特征金字塔结构,提高了模型对目标大小的鲁棒性。结果 本文模型在目前主流公开的食物图片数据集Food-101、ChineseFoodNet和Food-172上进行实验,分别获得了91.4%、82.8%、90.3%的Top-1正确率,与现有方法相比提高了1%~8%。结论 本文提出了一种多级卷积神经网络食物图片识别模型,可以自动定位食物图片区分度较大的区域,融合食物图片的全局与局部特征,实现了食物图片的细粒度识别,有效提高了食物图片的识别精度。实验结果表明,该模型在目前主流食物图片数据集上取得了最好的结果。