基于BoF模型的多特征融合纹理图像分类

针对特征词袋（BoF）模型缺乏空间和几何信息,对纹理图像内容表达不明显等问题,提出一种基于BoF模型的多特征融合纹理分类算法。将灰度梯度共生矩阵（GGCM）和尺度不变特征转换（SIFT）融合特征作为纹理图像的区域特征描述,通过动态权重鉴别能量分析进行最优参数特征选择,并用BoF量化纹理特征,使用支持向量机对图像进行训练和预测,得出分类结果。实验结果表明,本文算法对有旋转扭曲的纹理、边缘模糊纹理、有光照变化的纹理及杂乱纹理等均能取得较好的分类效果,相对于传统BoF模型及凹凸划分（CCP）方法等算法在UIUC纹理库上的分类正确率均有不同程度的提高,平均分类正确率分别提高12.8%和7.9%,说明本文算法针对纹理图像分类具有较高的精度和较好的鲁棒性。      

核空间聚类在图像纹理分类中的简化算法

模糊c均值聚类已广泛应用于模糊模式识别领域,但对于线性不可分数据并不适用.在核方法中通过将输入数据经过非线性映射投影到高维特征空间来解决非线性分类的问题.将传统的模糊c均值聚类算法应用于核空间中,对线性不可分的样本进行了核空间聚类的分类实验,得到了正确的分类结果.由于图像分类中分类样本(对应图像像素)数目庞大,造成了核空间聚类算法中特征距离的计算量过大.因此,在核空间聚类的基础上,提出了对图像先进行过分割,再对过分割的图像块进行核空间聚类的方法,大大降低了高维空间特征距离计算的运算成本,并取得了良好的分类效果.      

改进YOLOv5s的弱光水下生物目标检测算法

针对水下光学图像目标检测过程中由于水中光线衰弱严重、图像环境复杂和拍摄设备移动等造成的生物识别精度低的问题,提出了基于改进YOLOv5s的弱光水下生物目标实时检测算法YOLOv5s-underwater。针对弱光水下光线衰弱的问题,引入了限制对比度自适应直方图均衡(CLAHE)算法对输入图像进行预处理,解决了颜色失真和图像毛糙的问题。针对复杂的弱光水下图像环境,提出了快速空间金字塔池化(SPPF)模块,解决了水下物体区分度低和特征损失严重的问题。针对拍摄设备移动带来的场景和形态变化问题,提出了一种基于旋转窗口的SwinTransformer模块,提高了模型的泛化能力。针对水下小目标,修改了网络模型结构,提高了小目标的检测能力。仿真和实验结果表明：所提算法相较于YOLOv5s检测精度提高30.7%,证明了算法的有效性。     

基于物理模型的彩色图像理解算法

给出一个彩色图像理解算法。算法首先对图像上的颜色变化(由光照和物体颜色引起)进行分析与综合,然后分割图像。算法的基础是二色反射模型理论,该理论认为反射光的颜色是界面反射(高光颜色)和本体反射(物体颜色)的线性组合,这两种反射光在颜色空间的三维直方图中形成特定的聚类(点簇)。因此分析聚类的性质可帮助确定光照和物体的颜色,但是有意义的聚类的生成又以图像中物体区域的确定为前提。算法按照假设检验的策略,依据图像中的连通性和颜色空间中聚类的特征,完成彩色图像的分割,并产生对景物中所发生的光学过程的物理描述。     

一种基于FPGA的图像自动增强算法与实现

图像灰度分布不均匀是影响画质的重要因素，集中表现在对空间场景成像时，图像中存在大量过亮或过暗的区域.同时，算法的实时性和嵌入式平台的可实现性是设计实现空间应用载荷需要着重考虑的问题.针对上述问题，提出一种适合于FPGA实现的图像自动增强算法.算法以分段线性变换方法为基础，采用K均值聚类对视频图像的灰度直方图进行动态的区间划分；构建分段函数系数与直方图之间的定量关系，自动计算线性变换系数；采用高效的并行流水线结构设计，实现了基于FPGA的硬件处理系统.仿真和成像试验结果表明，该FPGA硬件系统实时性好，适应性强，针对不同的场景均取得较好的处理效果，具有广阔的应用前景.      

基于核方法的聚类算法及其应用

在分析核方法的核心概念基础上,提出了一种基于核方法的聚类算法.通常,传统聚类算法只在数据特征差异较大时才有效,当数据特征差异较小时,很难取得较好的聚类效果.引入核函数,将原始数据由数据空间映射到特征空间,在特征空间中进行聚类.核函数的非线性映射使得原始数据的特征更完整地显现出来,从而能够更客观准确地聚类.与传统聚类方法相比,该方法聚类结果更客观有效.以16组实际数据为例,将该方法应用于数据分类研究中,聚类结果表明了该方法的可行性和有效性,从而为数据分类提供了一种新的可行方法.      

基于图论分割的多光谱图像非监督分类方法

针对传统基于像素的多光谱遥感图像分类方法存在的"麻点"现象、采样成本高等问题,提出了一种基于图论分割的非监督分类方法,首先采用基于图论的分割算法,按局部邻近相似像素点分割成若干子区域,再以分割后子区域为基本单元,整体进行模糊 C均值聚类,最终实现对多光谱图像的非监督分类.实验证明,该方法结合了局部邻近像素点的相互关系以及相似区域的整体特征,有效解决了麻点问题,具有较高的分类精度和算法效率,降低了采样成本.     

一种迭代聚合的高分辨率网络Anchor-free目标检测方法

针对目前Anchor-free目标检测方法CenterNet（ObjectsasPoints）生成热力图不准确、检测精度不足的问题,提出了一种基于特征迭代聚合的高分辨率表征网络CenterNet-DHRNet。首先,引入高分辨率表征骨干网络,并用迭代聚合的方式对不同分辨率的特征图进行融合,提高网络的分辨率,有效减少图像在下采样过程中损失的空间语义信息。其次,使用高效通道注意力机制对高分辨率表征骨干网络的输出进行优化。最后,利用结合空洞卷积的空间金字塔池化操作增强网络对不同尺度物体的感受野。实验在PASCALVOC数据集和KITTI数据集上进行,结果表明:CenterNet-DHRNet精度更高,满足实时检测的性能要求,具有良好的鲁棒性。      

复杂动态场景下目标检测与分割算法

在动态场景等复杂条件下,往往难以对序列图像目标进行准确的检测与分割。根据序列图像中目标在复杂条件下的成像特点,提出了一种基于融合尺度不变特征变换(SIFT)流特征显著模型的动态场景目标检测与分割算法。通过对SIFT流算法表示运动特征信息的优势进行分析,并结合图像国际照明协会(CIE)Lab颜色空间的颜色和亮度特征信息,建立四维特征向量空间。利用改进的多尺度中心-环绕对比方法生成各特征通道的显著图并进行线性融合,建立序列图像的动态场景目标显著模型。最后利用均值漂移聚类算法和形态学处理实现对检测目标的精确分割。实验结果表明,相比传统检测与分割算法,该算法在动态背景与航拍等复杂场景下能够分割出更为完整的目标区域,具有良好的鲁棒性和高分割精度。      

基于MSER的无人机图像建筑区域提取

对建筑区域自动检测与提取是无人机(UAV,Unmanned Aerial Vehicle)图像处理的一项重要功能.在分析无人机成像特点和最大稳定极值区域(MSER,Maximum Stable Extremal Regions)算法对无人机侦察图像建筑区域检测的适用性基础上,提出了一种基于MSER的无人机侦察图像建筑区域提取算法.算法包含5步:无人机图像预处理,运用MSER算法分析计算图像稳定区域,通过计算稳定区域密度筛选建筑区域,进一步利用自适应K均值聚类算法对建筑区进行划分,最后采用Graham算法生成建筑区的边界从而实现了建筑区的自动提取.选取无人机实飞图像数据进行实验统计,本算法提取精度为92.25%;同时与基于Gabor变换的纹理特征、SIFT特征点的提取算法相比,建筑区域提取时间缩短,满足无人机实时应用需求.      

遥感图像的显著-概要特征提取与目标检测

针对巨幅遥感图像的目标检测问题,提出了一种基于显著-概要特征的遥感图像自动目标检测算法.采用滑动窗口将巨幅遥感图像划分为若干个小尺度的区域,针对各个小尺度分块图像,借鉴人类视觉生理功能特性之原理,提取其显著特征和概要特征,其中的显著特征代表了图像中的显著信息及显著区域空间分布和关联信息,概要特征可从整体上反映该区域的背景/目标关联信息.通过对分块区域图像的分类鉴别以实现目标检测.实验结果表明:此方法能以高可靠性和高精确度检测出巨幅遥感图像中的目标.     

基于超像素与多尺度残差U-Net相结合的遥感图像飞机检测方法

遥感图像中存在飞机很小、角度和位置不确定且背景复杂等问题,从遥感图像中检测飞机是一项重要且具有挑战性的任务,因此,提出一种基于超像素与多尺度残差U-Net(Multi-scale Residual U-Net,MSRU-Net)相结合的遥感图像飞机检测方法。首先对遥感图像进行超像素预分割,将位置相邻且像素特征相似的像素点组成若干个超像素,保持图像进一步分割的有效特征;然后构建多尺度残差U-Net,学习其多尺度判别特征。与传统的飞机检测方法相比,该方法用少量的超像素代替大量像素表达图像特征,降低了图像分割的复杂度,再利用MSRU-Net分割遥感超像素图像,有效检测不同尺度的飞机图像。在公共飞机遥感图像数据集上实验,结果表明,该方法能够有效的检测遥感图像不同尺度的飞机图像,检测精确率达到91.2 %。     

Advanced fractal approach for unsupervised classification of SAR images

Unsupervised classification of Synthetic Aperture Radar (SAR) images is the alternative approach when no or minimum apriori information about the image is available. Therefore, an attempt has been made to develop an unsupervised classification scheme for SAR images based on textural information in present paper. For extraction of textural features two properties are used viz. fractal dimension D and Moran’s I. Using these indices an algorithm is proposed for contextual classification of SAR images. The novelty of the algorithm is that it implements the textural information available in SAR image with the help of two texture measures viz. D and I. For estimation of D, the Two Dimensional Variation Method (2DVM) has been revised and implemented whose performance is compared with another method, i.e., Triangular Prism Surface Area Method (TPSAM). It is also necessary to check the classification accuracy for various window sizes and optimize the window size for best classification. This exercise has been carried out to know the effect of window size on classification accuracy. The algorithm is applied on four SAR images of Hardwar region, India and classification accuracy has been computed. A comparison of the proposed algorithm using both fractal dimension estimation methods with the K-Means algorithm is discussed. The maximum overall classification accuracy with K-Means comes to be 53.26% whereas overall classification accuracy with proposed algorithm is 66.16% for TPSAM and 61.26% for 2DVM.     

一种结合全局和局部相似性的小样本分割方法

针对小样本分割中如何提取支持图像和查询图像共性信息的问题,提出一种新的小样本分割模型,同时结合了全局相似性和局部相似性,实现了更具泛化能力的小样本分割。具体地,根据支持图像和查询图像全局特征和局部特征之间的相似性,提出了一种新型注意力谱生成器,进而实现查询图像的注意力谱生成和区域分割。所提注意力谱生成器包含2个级联模块:全局引导器和局部引导器。在全局引导器中,提出了一种新的基于指数函数的全局相似性度量,对查询图像特征和支持图像的全局特征进行关系建模,输出前景增强的查询图像特征。在局部引导器中,通过引入局部关系矩阵对支持图像特征和查询图像特征之间的局部相似性进行建模,得到与类别无关的注意力谱。在Pascal-5i数据集上做了大量的实验,在1-shot设定下mIoU达到了59.9%,5-shot设定下mIoU达到了61.9%,均优于现有方法。      

Crater detection via genetic search methods to reduce image features

Recent approaches to crater detection have been inspired by face detection’s use of gray-scale texture features. Using gray-scale texture features for supervised machine learning crater detection algorithms provides better classification of craters in planetary images than previous methods. When using Haar features it is typical to generate thousands of numerical values from each candidate crater image. This magnitude of image features to extract and consider can spell disaster when the application is an entire planetary surface. One solution is to reduce the number of features extracted and considered in order to increase accuracy as well as speed. Feature subset selection provides the operational classifiers with a concise and denoised set of features by reducing irrelevant and redundant features. Feature subset selection is known to be NP-hard. To provide an efficient suboptimal solution, four genetic algorithms are proposed to use greedy selection, weighted random selection, and simulated annealing to distinguish discriminate features from indiscriminate features. Inspired by analysis regarding the relationship between subset size and accuracy, a squeezing algorithm is presented to shrink the genetic algorithm’s chromosome cardinality during the genetic iterations. A significant increase in the classification performance of a Bayesian classifier in crater detection using image texture features is observed.     

Development and testing of a statistical texture model for land cover classification of the Black Sea region with MODIS imagery

A statistical model is proposed for analysis of the texture of land cover types for global and regional land cover classification by using texture features extracted by multiresolution image analysis techniques. It consists of four novel indices representing second-order texture, which are calculated after wavelet decomposition of an image and after texture extraction by a new approach that makes use of a four-pixel texture unit. The model was applied to four satellite images of the Black Sea region, obtained by Terra/MODIS and Aqua/MODIS at different spatial resolution. In single texture classification experiments, we used 15 subimages (50 × 50 pixels) of the selected classes of land covers that are present in the satellite images studied. These subimages were subjected to one-level and two-level decompositions by using orthonormal spline and Gabor-like spline wavelets. The texture indices were calculated and used as feature vectors in the supervised classification system with neural networks. The testing of the model was based on the use of two kinds of widely accepted statistical texture quantities: five texture features determined by the co-occurrence matrix (angular second moment, contrast, correlation, inverse difference moment, entropy), and four statistical texture features determined after the wavelet transformation (mean, standard deviation, energy, entropy). The supervised neural network classification was performed and the discrimination ability of the proposed texture indices was found comparable with that for the sets of five GLCM texture features and four wavelet-based texture features. The results obtained from the neural network classifier showed that the proposed texture model yielded an accuracy of 92.86% on average after orthonormal wavelet decomposition and 100% after Gabor-like wavelet decomposition for texture classification of the examined land cover types on satellite images.     

Detection of sub-kilometer craters in high resolution planetary images using shape and texture features

Counting craters is a paramount tool of planetary analysis because it provides relative dating of planetary surfaces. Dating surfaces with high spatial resolution requires counting a very large number of small, sub-kilometer size craters. Exhaustive manual surveys of such craters over extensive regions are impractical, sparking interest in designing crater detection algorithms (CDAs). As a part of our effort to design a CDA, which is robust and practical for planetary research analysis, we propose a crater detection approach that utilizes both shape and texture features to identify efficiently sub-kilometer craters in high resolution panchromatic images. First, a mathematical morphology-based shape analysis is used to identify regions in an image that may contain craters; only those regions - crater candidates - are the subject of further processing. Second, image texture features in combination with the boosting ensemble supervised learning algorithm are used to accurately classify previously identified candidates into craters and non-craters. The design of the proposed CDA is described and its performance is evaluated using a high resolution image of Mars for which sub-kilometer craters have been manually identified. The overall detection rate of the proposed CDA is 81%, the branching factor is 0.14, and the overall quality factor is 72%. This performance is a significant improvement over the previous CDA based exclusively on the shape features. The combination of performance level and computational efficiency offered by this CDA makes it attractive for practical application.     

用于飞行器视觉导航的地平线检测算法

结合航拍视频序列的特点与视觉导航系统的需求,提出了一种基于图像区域相似性和禁忌搜索算法的地平线拟合视觉导航算法.从航拍视频序列中提取单帧图像,并做下采样去噪等相应预处理;初始化禁忌搜索算法的初始解等参数,结合适配值函数,拟合图像中的地平线;用禁忌搜索算法的特赦准则和收敛准则评判拟合的准确性,直到适配值函数值满足收敛准则为止,得出最优解;分解出飞行器的横滚角和俯仰角,并由横滚角和俯仰角的大小决定下次参数校正的间隔,输出结果给控制系统.实验结果表明,与传统视觉导航方法和基于Hough变换的方法相比,此算法具有无参考、实时性好和抗噪性强等特点.