结合运动信息与表观特征的行人检测方法

提出一种结合运动信息与表观特征的行人检测方法.在对通过表观检测子获得的候选检测窗口执行分割验证的框架中,将运动信息融入到基于图像序列的对象分割算法中,通过获取更准确的分割结果来提高对候选检测窗口的检测准确率.该方法利用运动信息更新运动对象的前景/背景分布模型,将颜色信息间接地融入行人检测中,并通过形状特征表现出来,与行人表观检测子形成互补的特性,获得更好的检测结果.上述结论在CAVIAR视频以及行人检测视频中得到了实验验证.     

基于似物性和空时协方差特征的行人检测算法

针对行人检测算法中缺少空时信息融合、检测区域过大等问题,提出了一种联合似物性检测和基于通道协方差信息的改进算法。该算法首先对图像进行二进制梯度归一化的似物性检测,并形成行人检测候选区域,缩小检测区域;然后提取待测目标的空间和时间特征;最后基于协方差信息构造一种融合空时特征的检测器,以提高检测精度。在公开的数据集INRIA和Caltech上的实验结果表明:该算法的性能优于目前主流的行人检测算法。     

基于DCT变换的多通道特征的行人检测

在行人检测中,针对目前多通道检测算法特征利用不充分的问题,提出一种基于DCT变换的多通道特征级联的行人检测算法。通过一种2层卷积网络模型将图像信息DCT变换后的数据进行整理,形成新的频域通道特征,该通道能描述行人的复杂纹理特征。结合梯度方向直方图特征、颜色空间特征和DCT频域特征,基于Adaboost算法训练了低开销的多通道特征行人检测器。在典型的公开行人库上的实验结果表明,该方法能提高检测的性能,在较低误检率时效果更加显著。     

深度学习行人检测方法综述

行人检测技术在智能交通系统、智能安防监控和智能机器人等领域均表现出了极高的应用价值,已经成为计算机视觉领域的重要研究方向之一。得益于深度学习的飞速发展,基于深度卷积神经网络的通用目标检测模型不断拓展应用到行人检测领域,并取得了良好的性能。但是由于行人目标内在的特殊性和复杂性,特别是考虑到复杂场景下的行人遮挡和尺度变化等问题,基于深度学习的行人检测方法也面临着精度及效率的严峻挑战。本文针对上述问题,以基于深度学习的行人检测技术为研究对象,在充分调研文献的基础上,分别从基于锚点框、基于无锚点框以及通用技术改进（例如损失函数改进、非极大值抑制方法等）3个角度,对行人检测算法进行详细划分,并针对性地选取具有代表性的方法进行详细结合和对比分析。本文总结了当前行人检测领域的通用数据集,从数据构成角度分析各数据集应用场景。同时讨论了各类算法在不同数据集上的性能表现,对比分析各算法在不同数据集中的优劣。最后,对行人检测中待解决的问题与未来的研究方法做出预测和展望。如何缓解遮挡导致的特征缺失问题、如何应对单一视角下尺度变化问题、如何提高检测器效率以及如何有效利用多模态信息提高行人检测精度,均是值得进一步研究的方向。     

Pedestrian detection with unsupervised multispectral feature learning using deep neural networks

Multispectral pedestrian detection is an important functionality in various computer vision applications such as robot sensing, security surveillance, and autonomous driving. In this paper, our motivation is to automatically adapt a generic pedestrian detector trained in a visible source domain to a new multispectral target domain without any manual annotation efforts. For this purpose, we present an auto-annotation framework to iteratively label pedestrian instances in visible and thermal channels by leveraging the complementary information of multispectral data. A distinct target is temporally tracked through image sequences to generate more confident labels. The predicted pedestrians in two individual channels are merged through a label fusion scheme to generate multispectral pedestrian annotations. The obtained annotations are then fed to a two-stream region proposal network (TS-RPN) to learn the multispectral features on both visible and thermal images for robust pedestrian detection. Experimental results on KAIST multispectral dataset show that our proposed unsupervised approach using auto-annotated training data can achieve performance comparable to state-of-the-art deep neural networks (DNNs) based pedestrian detectors trained using manual labels.     

基于深度卷积神经网络特征层融合的小尺度行人检测

针对目前行人检测算法计算量过大和对小尺度行人检测精度不高的问题,提出了一种基于深度卷积神经网络特征层融合的小尺度行人检测方法,设计了一种包含9个卷积层的深度神经网络架构.在进行行人检测时,首先,对输入图像进行分块预处理操作,避免损失原始图像的视觉信息;然后,将网络不同层的卷积特征进行融合,提升行人特征的区分能力和表达能力,进而提升行人检测的精度,在保证检测精度的同时有效降低网络的复杂度.在INRIA、Caltech等公共行人数据集上的实验结果表明,所提出的行人检测方法能够有效检测小尺度的行人,且网络架构的参数量更少,检测速度更快,能得到更高精度的行人检测结果.     

基于运动图像序列的异常行为检测*

针对公共重点区域的智能监视问题,研究了一种基于运动历史图像（motion history image,MHI）的行人异常行为检测方法。利用运动图像序列得到的MHI获取视频帧中运动目标的运动方向,由运动方向的变化分类确定人体运动模式和行为是否异常,同时给出相应的实验结果。结果表明,该方法实现简单,具有较好的实时性与鲁棒性,可以作为实时监控系统中异常行为检测的有效方法。     

Comparison of edge detection algorithms using a structure frommotion task

This paper presents an evaluation of edge detector performance. We use the task of structure from motion (SFM) as a "black box" through which to evaluate the performance of edge detection algorithms. Edge detector goodness is measured by how accurately the SFM could recover the known structure and motion from the edge detection of the image sequences. We use a variety of real image sequences with ground truth to evaluate eight different edge detectors from the literature. Our results suggest that ratings of edge detector performance based on pixel-level metrics and on the SFM are well correlated and that detectors such as the Canny detector and Heitger detector offer the best performance.     

Multi-modal time-of-flight based fire detection

This paper proposes two novel time-of-flight based fire detection methods for indoor and outdoor fire detection. The indoor detector is based on the depth and amplitude image of a time-of-flight camera. Using this multi-modal information, flames can be detected very accurately by fast changing depth and amplitude disorder detection. In order to detect the fast changing depth, depth differences between consecutive frames are accumulated over time. Regions which have multiple pixels with a high accumulated depth difference are labeled as candidate flame regions. Simultaneously, the amplitude disorder is also investigated. Regions with high accumulative amplitude differences and high values in all detail images of the amplitude image its discrete wavelet transform, are also labeled as candidate flame regions. Finally, if one of the depth and amplitude candidate flame regions overlap, fire alarm is given. The outdoor detector, on the other hand, only differs from the indoor detector in one of its multi-modal inputs. As depth maps are unreliable in outdoor environments, the outdoor detector uses a visual flame detector instead of the fast changing depth detection. Experiments show that the proposed detectors have an average flame detection rate of 94% with no false positive detections.     

基于部位检测和子结构组合的行人检测方法

提出了一种基于部位检测和子结构组合的、可用于辅助驾驶或视频监控系统中行人检测的方法.首先使用头部分类器在整幅图像中检测,得到感兴趣区域;然后在每个感兴趣区域内使用头部、躯干、腿部以及左臂和右臂5个人体部位检测器分别检测并使用基于子结构的检测组合方法对部位检测结果进行组合,以得到最终结果.在不同数据库上的实验结果表明,本方法可以有效地用于移动或静止摄像机所拍摄的视频图像中的多姿态及部分遮挡的行人检测.