一种视频监控中基于航迹的运动小目标检测算法

针对视频监控中运动小目标难以检测的问题,该文提出一种基于航迹的检测算法。首先,为了降低检测漏警率,提出区域纹理特征与差值概率融合的自适应前景提取方法;其次,为了降低检测虚警率,设计航迹关联的概率计算模型以建立疑似目标在视频帧间的关联,并设置双门限以区分疑似目标中的真实目标与虚假目标。实验结果表明,与多种经典算法相比,该算法能对定量范围内的运动小目标以更低的漏警率和虚警率实施准确检测。     

引入自上向下特征融合的小目标检测算法

针对原始SSD(Single Shot Multibox Detector)算法未充分利用各特征层之间关系导致浅层特征层缺乏小目标语义信息的问题,为了提高对小目标的检测能力,提出了一种结合PANet多尺度特征融合网络和自上向下特征融合路径的TTB-SSD(Top to Bottom SSD)改进算法。首先,使用PANet多尺度特征融合网络对特征进行反复提取,从而获得丰富的多尺度语义信息;然后,使用一种深层特征融合模块将浅层特征层的空间信息传递到深层特征层,进而更准确地对小目标进行定位;最后,为了增强浅层特征层的语义信息,构造了自上向下的特征融合路径,从而强化浅层对小目标检测的准确率。实验结果表明,在PASCAL VOC2007测试集检测的mAP(Mean Average Precision)值达到80.5%,对目标的mAP较原始SSD提高了5.7%,证明了该算法对小目标检测的有效性。     

基于深度学习的小目标检测研究与应用综述

目标检测在基于传统手工特征及深度学习算法上已经取得较大发展,然而针对小目标检测的研究近几年才开始出现,研究成果较少,且大都是在已有目标检测算法基础上进行改进,以提高小目标检测的检测精度.小目标像素点少,本身携带的特征少,多次下采样后就更难进行特征提取,因而小目标检测面临极大挑战.小目标检测在自动驾驶、遥感图像检测、刑侦等领域都有广泛应用需求,对于小目标检测技术的研究有重要的实用价值.本文对小目标检测的现有研究成果进行了详细综述.首先,将现有算法按照检测需要的阶段数分为一阶段、二阶段、多阶段,描述了RetinaNet、CornerNet-Lite、特征金字塔网络（Feature Pyramid Network,FPN）等算法的原理并进行了对比分析.其次,本文描述了小目标检测技术在不同领域的应用情况,并汇总了MS COCO、PASCAL VOC、DOTA、KITTI等数据集及算法性能评价指标.最后,总结了小目标检测面临的挑战,并展望了未来的研究方向.     

一种基于多尺度特征融合的目标检测算法

基于深度学习的目标检测器RetinaNet和Libra RetinaNet均是使用特征金字塔网络融合多尺度特征,但上述两个检测器存在特征融合不充分的问题。鉴于此,提出一种多尺度特征融合算法。该算法是在Libra RetinaNet的基础上进一步扩展,通过建立两条自底向上的路径构建两个独立的特征融合模块,并将两个模块产生的结果与原始预测特征融合,以此提高检测器的精度。将多尺度特征融合模块与Libra RetinaNet结合构建目标检测器并在不同的数据集上进行实验。实验结果表明,与Libra RetinaNet检测器相比,加入模块后的检测器在PASCAL VOC数据集和MSCOCO数据集上的平均精度分别提高2.2个百分点和1.3个百分点。     

基于场景感知的运动目标检测方法

背景减除法是一种主要的运动目标检测框架,但在复杂环境中构建一种初始模型建立周期短、可靠性高、鲁棒性好的模型仍是一大难题.本文从场景感知的角度出发,在背景减除框架的基础上提出一种目标检测方法.该方法根据前两帧中稳定的结构信息感知背景中潜在的前景区域,在第二帧建立初始模型时利用最近邻域背景像素点代替可能的前景像素点,提高了初始模型可靠性;结合颜色信息和二进制特征提出了像素点二级分类判决机制,并通过感知像素点邻域内的纹理复杂度自适应调整局部判决阈值和更新频率;在模型更新阶段提出处理误判的反馈机制.在公开视频序列上同几种流行检测算法的实验对比结果证明了本文算法的有效性和优越性.     

一种快速的图像小目标检测算法

针对图像小目标检测问题进行了研究,在分析现有研究成果的基础上,充分利用了遗传算法动态多点搜索的快速性,以及分形特征应用于目标检测的稳定性,提出一种基于遗传算法与分形特征的快速图像小目标检测算法,并对算法进行了仿真。仿真结果表明该算法有效地提高了图像小目标检测的速度。     

面向无人机影像的目标特征跟踪方法研究

提出了一种面向无人机视频影像的目标特征跟踪模型,首先根据目标检测结果在目标感兴趣区域中采用Shi-Tomasi算法进行角点检测,提取目标特征角点并在跟踪过程中对特征角点进行实时更新以保证角点数量满足跟踪条件,最后通过金字塔Lucas-Kanade算法对目标特征进行跟踪。实验结果表明,该模型鲁棒性强且易于实现,在获得良好跟踪效果的同时可满足实时处理要求。     

基于多层次互补特征的通用目标检测模型

针对实际场景中多类目标检测问题,该文提出了一种基于多层次特征表示和异质互补描述子的通用目标检测模型。该模型采用基于组件的目标描述思想,提取目标不同层次的互补特征,并将其统一到条件随机场(CRF)框架中。目标中单个组件及其局部特征对应CRF的一元节点,组件之间的几何空间结构特征则体现在节点之间的两两连接关系上。通过引入节点支持向量机(SVM)分类器和边缘拓扑结构学习,极大提高了模型的鉴别能力和灵活性。在UIUC多尺度数据集和PASCAL VOC 2007数据集上测试结果表明,该文模型不仅能有效描述多类复杂目标,还能较好地解决姿态、尺度、光照变化及局部遮挡情况下的目标检测问题。     

基于细化多尺度深度特征的目标检测网络

现有深度卷积神经网络中感受野尺度单一,无法适应目标的尺度变化和边界形变,故此本文提出了一种提取并融合多尺度特征的目标检测网络.该网络通过减少池化并在网络底层加入空间加信道压缩激励模块来突出可利用的细节信息,生成高质量的特征图;此外,在深层网络中加入可变多尺度特征融合模块,该模块具有多种尺度的感受野并可根据物体边界预测采样位置,最后通过融合多尺度特征使网络具有更强的特征表达能力并且对不同尺度实例及其边界信息更具鲁棒性.实验证明,本文结构实现了比原有结构更高的平均精度,与目前主流目标检测算法相比也具有一定优势.     

基于梯度双面互补特性的级联快速目标检测

针对目标检测中精度和速度难以兼顾的问题,借助视觉注意理论中的目标感知与识别机制,分析目标描述中梯度幅值与梯度方向信息之间具有的互补性,提出了基于两层级联梯度特征的快速目标检测模型,可有效描述类无关和类相关检测器.一方面,采用梯度幅值特征,从滑动窗口采样中获得候选目标提议,大幅降低了验证窗口的数量,确保检测速度,另一方面,利用级联方式学习训练多个子检测器,可更好实现不同尺度变化下的目标检测精度.PASCAL数据集上的实验结果,解释了级联梯度特征对目标结构描述的有效性,表明了该文方法在与现有先进方法的检测精度相当的前提下,可极大提升检测速度.     

基于模板匹配的快速少纹理目标识别算法

针对传统的基于模板匹配算法通常考虑模板的整体性往往遇到计算性能复杂的问题。文中提出一种基于二进制方向压缩映射和局部特征加权的快速少纹理目标识别方法。根据目标边缘点的量化梯度方向,利用二进制方向压缩映射方法对目标模型进行特征描述,快速提取出目标候选位置及其对应的尺度、角度信息;在检测出目标候选位置后,再利用局部特征加权方法建立新的模板特征,对目标候选位置计算新的相似度从而确定目标最终姿态。实验结果表明,文中算法与其他具有代表性的算法相比具有更好的识别率,并且识别时间大幅降低。     

融合视觉机制和多尺度特征的小目标检测算法

针对SSD(Single Shot MultiBox Detector)目标检测算法对小目标检测能力不足的问题,提出一种引入视觉机制和多尺度语义信息融合的VFF-SSD(Vision Feature Fusion SSD)改进算法。为了增大浅层网络的感受野提高特征提取能力,首先在SSD浅层特征层中加入视觉机制,然后利用改进PANet(Path Aggregation Network)多尺度特征融合网络与深层特征增强网络得到新的特征层,旨在增强浅层网络的语义信息并加强深层特征的特征表达能力,最后应用注意力机制模块提高对重要信息的学习能力。实验结果表明,在PASCAL VOC2007测试集检测的mAP(Mean Average Precision)值达到81.1%,对数据集中小目标的mAP值较原SSD提高了6.6%。     

Global and local information aggregation network for edge-aware salient object detection

Aggregation of local and global contextual information by exploiting multi-level features in a fully convolutional network is a challenge for the pixel-wise salient object detection task. Most existing methods still suffer from inaccurate salient regions and blurry boundaries. In this paper, we propose a novel edge-aware global and local information aggregation network (GLNet) to fully exploit the integration of side-output local features and global contextual information and utilization of contour information of salient objects. The global guidance module (GGM) is proposed to learn discriminative multi-level information with the direct guidance of global semantic knowledge for more accurate saliency prediction. Specifically, the GGM consists of two key components, where the global feature discrimination module exploits the inter-channel relationship of global semantic features to boost representation power, and the local feature discrimination module enables different side-output local features to selectively learn informative locations by fusing with global attentive features. Besides, we propose an edge-aware aggregation module (EAM) to employ the correlation between salient edge information and salient object information for generating estimated saliency maps with explicit boundaries. We evaluate our proposed GLNet on six widely-used saliency detection benchmark datasets by comparing with 17 state-of-the-art methods. Experimental results show the effectiveness and superiority of our proposed method on all the six benchmark datasets.     

FD-SSD: An improved SSD object detection algorithm based on feature fusion and dilated convolution

Objects that occupy a small portion of an image or a frame contain fewer pixels and contains less information. This makes small object detection a challenging task in computer vision. In this paper, an improved Single Shot multi-box Detector based on feature fusion and dilated convolution (FD-SSD) is proposed to solve the problem that small objects are difficult to detect. The proposed network uses VGG-16 as the backbone network, which mainly includes a multi-layer feature fusion module and a multi-branch residual dilated convolution module. In the multi-layer feature fusion module, the last two layers of the feature map are up-sampled, and then they are concatenated at the channel level with the shallow feature map to enhance the semantic information of the shallow feature map. In the multi-branch residual dilated convolution module, three dilated convolutions with different dilated ratios based on the residual network are combined to obtain the multi-scale context information of the feature without losing the original resolution of the feature map. In addition, deformable convolution is added to each detection layer to better adapt to the shape of small objects. The proposed FD-SSD achieved 79.1% mAP and 29.7% mAP on PASCAL VOC2007 dataset and MS COCO dataset respectively. Experimental results show that FD-SSD can effectively improve the utilization of multi-scale information of small objects, thus significantly improve the effect of the small object detection.     

运动序列中动目标检测的稳健性方法

提出一种运动序列中动目标检测的稳健性方法。用尺度不变特征变换(SIFT)算法生成特征描述符,基于最近邻距离比(NNDR)进行初始匹配,增加对称性约束以获得稳健的匹配点集。随机抽样一致集算法(RANSAC)用于分离背景和目标对应特征点,实现背景运动的稳健性估计。背景补偿后,相邻帧差分和数学形态学方法实现动目标的分割。真实运动序列的实验结果表明,该算法能够获得稳健的匹配点对,检测出运动目标。     

CANet: Context-aware Aggregation Network for Salient Object Detection of Surface Defects

Surface defect detection has become more and more important in the industrial manufacture and engineering application in recent years. However, due to the lack of overall perception and interaction among features layers, lots of computer vision-based detection methods cannot grab the complete details of defects when dealing with complex scenes, such as low contrast and irregular shape. Therefore, in this paper, we propose a Context-aware Aggregation Network (CANet) to accurately pop-out the defects, where we focus on the mining of context cues and the fusion of multiple context features. To be specific, embarking on the multi-level deep features extracted by encoder, we first deploy a sufficient exploration to dig the context information by deploying the weighted convolution pyramid (WCP) module, which extracts multi-scale context features, transfers the information flow between different resolution features, and fuses the features with same resolution. By this way, we can obtain the effective context pyramid features. Then, the decoder deploys the weighted context attention (WCA) module to filter the irrelevant information from context features and employs the cascaded fusion structure (CFS) to aggregate the multiple context cues in a hierarchical way. Following this way, the generated high-quality saliency maps can highlight the defects accurately and completely. Extensive experiments are performed on four public datasets, and the results firmly prove the effectiveness and superiority of the proposed CANet under different evaluation metrics.     

基于深度学习的目标检测研究综述

目标检测是计算机视觉领域内的热点课题,在机器人导航、智能视频监控及航天航空等领域都有广泛的应用.本文首先综述了目标检测的研究背景、意义及难点,接着对基于深度学习目标检测算法的两大类进行综述,即基于候选区域和基于回归算法.对于第一类算法,先介绍了基于区域的卷积神经网络（Region with Convolutional Neural Network,R-CNN）系列算法,然后从四个维度综述了研究者在R-CNN系列算法基础上所做的研究：对特征提取网络的改进研究、对感兴趣区域池化层的改进研究、对区域提取网络的改进研究、对非极大值抑制算法的改进研究.对第二类算法分为YOLO（You Only Look Once）系列、SSD（Single Shot multibox Detector）算法及其改进研究进行综述.最后根据当前目标检测算法在发展更高效合理的检测框架的趋势下,展望了目标检测算法未来在无监督和未知类别物体检测方向的研究热点.     

Nanomesh-YOLO: Intelligent Colorimetry E-Skin Based on Nanomesh and Deep Learning Object Detection Algorithm

Perspiration is an important physiological process that maintains thermal homeostasis and water–salt balance. However, the collection and analysis of perspiration currently rely on microfluidic technology and colorimetric assays. The complexity and high cost of fabricating microfluidic channels and the insecurity of chemical reagents for color reactions should be optimized. In this work, a colorimetry electronic skin (e-skin) for intelligent perspiration monitoring has been realized. The colorimetry e-skin system consists of the polyurethane (PU) nanomesh and the object detection algorithm You Only Look Once version 3 (YOLOv3). Due to the 44% porosity of the PU nanomesh and capillary action, the low-cost PU nanomesh (<1 cent) can be used as the colorimetric indicator. The volume of the PU nanomesh expands to 362.37% as a result of perspiration being absorbed and changes the optical transmittance (up to 277.78%). A finite element model based on capillary action has been proposed to explain the change in optical transmittance. Finally, a database containing 735 images has been built, and the object detection algorithm YOLOv3 is used to analyze the perspiration absorbed by the PU nanomesh. The detection results can identify the perspiration volume with a high accuracy of 97%. These results show that this work has great potential in healthcare field.     

基于深度帧差卷积神经网络的运动目标检测方法研究

复杂场景中的运动目标检测是计算机视觉领域的重要问题,其检测准确度仍然是一大挑战.本文提出并设计了一种用于复杂场景中运动目标检测的深度帧差卷积神经网络（Deep Difference Convolutional Neural Network,DFDCNN）.DFDCNN由DifferenceNet和AppearanceNet组成,不需要后处理就可以预测分割前景像素.DifferenceNet具有孪生Encoder-Decoder结构,用于学习两个连续帧之间的变化,从输入（t帧和t+1帧）中获取时序信息;AppearanceNet用于从输入（t帧）中提取空间信息,并与时序信息融合;同时,通过多尺度特征图融合和逐步上采样来保留多尺度空间信息,以提高网络对小目标的敏感性.在公开标准数据集CDnet2014和I2R上的实验结果表明：DFDCNN不仅在动态背景、光照变化和阴影存在的复杂场景中具有更好的检测性能,而且在小目标存在的场景中也具有较好的检测效果.