一种基于立体视觉的运动目标检测算法

在目标检测中采用立体视觉方法。首先对立体图像对进行匹配求取场景的视差图,再运用基于视差的背景差分法获得含有运动目标的前景区域,最后根据前景区域的视差和位置分布准确定位各运动目标。立体视觉方法有效解决了单目视觉检测方法中的一些难点问题,可以克服光线的变化和阴影干扰对目标检测带来的影响,在多个目标发生部分遮挡时仍能正确区分各运动目标。     

Change detection by probabilistic segmentation from monocular view

We present a method for foreground/background video segmentation (change detection) in real-time that can be used, in applications such as background subtraction or analysis of surveillance cameras. Our approach implements a probabilistic segmentation based on the Quadratic Markov Measure Field models. This framework regularizes the likelihood of each pixel belonging to each one of the classes (background or foreground). We propose a new likelihood that takes into account two cases: the first one is when the background is static and the foreground might be static or moving (Static Background Subtraction), the second one is when the background is unstable and the foreground is moving (Unstable Background Subtraction). Moreover, our likelihood is robust to illumination changes, cast shadows and camouflage situations. We implement a parallel version of our algorithm in CUDA using a NVIDIA Graphics Processing Unit in order to fulfill real-time execution requirements.     

Background subtraction model based on color and depth cues

Background subtraction consists of segmenting objects in movement in a video captured by a static camera. This is typically performed using color information, but it leads to wrong estimations due to perspective and illumination issues. We show that multimodal approaches based on the integrated use of color and depth cues produce more accurate and robust results than using either data source independently. Depth is less affected by issues such as shadows or foreground objects similar to background. However, objects close to the background may not be detected when using only range information, being color information complementary in those cases. We propose an extension of a well-known background subtraction technique which fuses range and color information, as well as a post-processing mask fusion stage to get the best of each feature. We have evaluated the method proposed using a well-defined dataset and different disparity estimation algorithms, showing the benefits of our method for fusion color and depth cues.     

基于光子映射的辐照度缓存算法

提出了一种全局光照计算方法,结合了两个知名的技术,光子映射和辐照度缓存.光子映射具有视点无关的优势,辐照度缓存可以快速计算间接光照,但后者是视点相关的,为了使光照缓存记录覆盖整个场景,辐照度缓存算法需要手动设置很多相机.利用这两种技术的各自优势,通过光子图来计算改进后的视点无关的辐照度缓存算法,实现了快速而准确的全局光...     

Pedestrian detection by means of far-infrared stereo vision

This article presents a stereo system for the detection of pedestrians using far-infrared cameras. Since pedestrian detection in far-infrared images can be difficult in some environmental conditions, the system exploits three different detection approaches: warm area detection, edge-based detection, and disparity computation. A final validation process is performed using head morphological and thermal characteristics. Currently, neither temporal correlation, nor motion cues are used in this processing.The developed system has been implemented on an experimental vehicle equipped with two infrared cameras and preliminarily tested in different situations.     

Human segmentation by geometrically fusing visible-light and thermal imageries

From depth sensors to thermal cameras, the increased availability of camera sensors beyond the visible spectrum has created many exciting applications. Most of these applications require combining information from these hyperspectral cameras with a regular RGB camera. Information fusion from multiple heterogeneous cameras can be a very complex problem. They can be fused at different levels from pixel to voxel or even semantic objects, with large variations in accuracy, communication, and computation costs. In this paper, we propose a system for robust segmentation of human figures in video sequences by fusing visible-light and thermal imageries. Our system focuses on the geometric transformation between visual blobs corresponding to human figures observed at both cameras. This approach provides the most reliable fusion at the expense of high computation and communication costs. To reduce the computational complexity of the geometric fusion, an efficient calibration procedure is first applied to rectify the two camera views without the complex procedure of estimating the intrinsic parameters of the cameras. To geometrically register different blobs at the pixel level, a blob-to-blob homography in the rectified domain is then computed in real-time by estimating the disparity for each blob-pair. Precise segmentation is finally achieved using a two-tier tracking algorithm and a unified background model. Our experimental results show that our proposed system provides significant improvements over existing schemes under various conditions.     

一种改进的自适应高斯混合模型实时运动目标检测算法

高斯混合模型被广泛应用于摄像机静止条件下运动目标检测的背景建模。针对传统高斯混合模型中对光照变化适应性差及学习率单一等问题, 提出了一种光照变化检测及学习率更新的方法, 以达到自适应更新背景模型的目的。提出利用颜色直方图匹配算法, 通过引入光照变化因子以及模型参数更新计数器对学习率进行自适应的调整, 并通过对描述模型分量个数的自适应选择减少了计算时间, 增强了系统的实时性。实验结果表明, 该方法能快速有效地适应场景的变化, 比传统高斯混合模型具有更好的鲁棒性与稳定性。     

Scalable hardware architecture for disparity map computation and object location in real-time

We present the disparity map computation core of a hardware system for isolating foreground objects in stereoscopic video streams. The operation is based on the computation of dense disparity maps using block-matching algorithms and two well-known metrics: sum of absolute differences and Census transform. Two sets of disparity maps are computed by taking each of the images as reference so that a consistency check can be performed to identify occluded pixels and eliminate spurious foreground pixels. Taking advantage of parallelism, the proposed architecture is highly scalable and provides numerous degrees of adjustment to different application needs, performance levels and resource usage. A version of the system for 640 × 480 images and a maximum disparity of 135 pixels was implemented in a system based on a Xilinx Virtex II-Pro FPGA and two cameras with a frame rate of 25 fps (less than the maximum supported frame rate of 40 fps on this platform). Implementation of the same system on a Virtex-5 FPGA is estimated to achieve 80 fps, while a version with increased parallelism is estimated to run at 140 fps (which corresponds to the calculation of more than 5.9 × 10⁹ disparity-pixels per second).     

Detection of moving objects with a moving camera using non-panoramic background model

This paper presents a fast and reliable method for moving object detection with moving cameras (including pan–tilt–zoom and hand-held cameras). Instead of building large panoramic background model as conventional approaches, we construct a small-size background model, whose size is the same as input frame, to decrease computation time and memory storage without loss of detection performance. The small-size background model is built by the proposed single spatio-temporal distributed Gaussian model and this can solve false detection results arising from registration error and background adaptation problem in moving background. More than the proposed background model based on spatial and temporal information, several pre- and post-processing methods are adopted and organized systematically to enhance the detection performances. We evaluate the proposed method with several video sequences under difficult conditions, such as illumination change, large zoom variation, and fast camera movement, and present outperforming detection results of our algorithm with fast computation time.     

Efficient illumination compensation techniques for text images

With the great advantages of digitization, more and more documents are being transformed into digital representations. Most content digitization of documents is performed by scanners or digital cameras. However, the transformation might degrade the image quality caused by lighting variations, i.e. uneven illumination distribution. In this paper we describe a new approach for text images to compensate uneven illumination distribution with a high degree of text recognition. Our proposed scheme is implemented by enhancing the contrast of the scanned documents, and then generating an edge map from the contrast-enhanced image for locating text area. With the information of the text location, a light distribution image (background) is created to assist the producing of the final light balanced image. Simulation results demonstrate that our approach is superior to the previous works of Hsia et al. (2005, 2006).     

Automatic illumination control for an endoscopy sensor

Endoscopy and small surgery can be performed with the support of small cameras. These cameras need to be minimal to reduce intrusion, require fast response and automatic illumination. The current work describes the world's smallest camera for endoscopy, NanEye, and the dynamic control algorithm for ROI developed for the illumination LED source coupled to the camera head that allows adjusting light, gain and exposure time. The obtained results show that it is capable of achieving correction speeds under 1 s while maintaining a static error below 3% relative to the total number of pixels of the image. The result of this work will allow the integration of high brightness LED sources of millimeter size on a very small camera enabling its use in endoscopic surgical robotic operations or minimal invasive surgery and NanEye is expected to be supplied at low prices for single time use.     

面向复杂背景中彩色立体图像的障碍物检测

目标识别技术作为图像理解领域的核心研究内容之一,一直以来是文献中讨论的热门话题.随着计算机科学的发展以及彩色成像设备的普及,越来越多的彩色立体图像在实际生产、生活中得到应用,面向彩色立体图像的目标识别技术逐渐成为吸引学者注意力的新方向.本文基于立体视觉信息,对安装在行进车辆上的两个摄像机在一段时间内获取的立体视频信息进行分析,提取车辆前方可能对行车安全造成威胁的障碍物.在V视差理论的基础上引入两个改进算法首先对原始立体图像进行数据净化和前景/背景分割,使后续算法不受复杂、多变背景的干扰,同时减少数据处理的冗余.然后使用绝对值差(SAD)[9]法直接计算彩色立体图像的浓密视差图,而无需彩色到灰度图像的转换.实验证明,本文提出的改进算法针对实际立体图像可以准确、快速地提取障碍物信息.     

Object tracking across non-overlapping views by learning inter-camera transfer models

In this paper, we introduce a novel algorithm to solve the problem of object tracking across multiple non-overlapping cameras by learning inter-camera transfer models. The transfer models are divided into two parts according to different kinds of cues, i.e. spatio-temporal cues and appearance cues. To learn spatio-temporal transfer models across cameras, an unsupervised topology recovering approach based on N-neighbor accumulated cross-correlations is proposed, which estimates the topology of a non-overlapping multi-camera network. Different from previous methods, the proposed topology recovering method can deal with large amounts of data without considering the size of time window. To learn inter-camera appearance transfer models, a color transfer method is used to model the changes of color characteristics across cameras, which has an advantage of low requirements to training samples, making update efficient when illumination conditions change. The experiments are performed on different datasets. Experimental results demonstrate the effectiveness of the proposed algorithm.     

Shell radiance texture functions

The appearance of an inhomogeneous translucent material depends substantially on its volumetric variations and their effects upon subsurface scattering. For efficient rendering that accounts for both surface mesostructures and volumetric variations of such materials, shell texture functions have precomputed irradiance within a volume with respect to incoming illumination, but even with this irradiance data a fair amount of runtime computation is still required. Rather than precompute volume irradiance, we introduce the shell radiance texture function (SRTF), which relates incoming illumination more directly to outgoing surface radiance by representing a set of subsurface transport components from which surface radiance can be calculated without ray marching or runtime evaluation of dipole diffusion. Using this precomputed SRTF information, inhomogeneous objects can be rendered in real time with distant local lighting or global lighting.     

前背景骨架区域随机点对策略驱动下的车标识别方法

目的 现有的车标识别算法均为各种经典的图像特征算子结合不同的分类器组合而成,均未分析车标图像的结构特点。综合考虑车标图像的灰度特征和结构特征,提出了一种前背景骨架区域随机点对策略驱动下的车标识别方法。方法 本文算法将标准车标图像分为前景区域和背景区域,分别提取前、背景的骨架区域,在其中进行随机取点,形成点对,通过进行点对的有效性判断,提取能表示车标的点对特征。点对特征表示两点周围局部区域的相似关系,反映了实际车标成像过程中车标图案部分与背景部分的灰度明暗关系。结果 在卡口系统截取的19 044张车标图像上进行实验,结果表明,与其他仅基于灰度特征的识别方法相比,本文提出的点对特征识别方法具有更好的识别效果,识别率达到了95.7%。在弱光照条件下,本文算法的识别算法效果同样优于其他仅基于灰度特征的识别方法,识别率达到了87.2%。结论 本文提出的前背景骨架区域随机点对策略驱动下的车标识别方法,结合了车标图像的灰度特征和结构特征,在进行车标的描述上具有独特性和排他性,有效地提高了车标的识别率,尤其是在弱光照条件下,本文方法具有更强的鲁棒性。     

Photon‐driven Irradiance Cache

We describe a global illumination method combining two well known techniques: photon mapping and irradiance caching. The photon mapping method has the advantage of being view independent but requires a costly additional rendering pass, called final gathering. As for irradiance caching, it is view‐dependent, irradiance is only computed and cached on surfaces of the scene as viewed by a single camera. To compute records covering the entire scene, the irradiance caching method has to be run for many cameras, which takes a long time and is a tedious task since the user has to place the needed cameras manually. Our method exploits the advantages of these two methods and avoids any intervention of the user. It computes a refined, view‐independent irradiance cache from a photon map. The global illumination solution is then rendered interactively using radiance cache splatting.     

Continuous localization and mapping of a pan–tilt–zoom camera for wide area tracking

Pan–tilt–zoom (PTZ) cameras are well suited for object identification and recognition in far-field scenes. However, the effective use of PTZ cameras is complicated by the fact that a continuous online camera calibration is needed and the absolute pan, tilt and zoom values provided by the camera actuators cannot be used because they are not synchronized with the video stream. So, accurate calibration must be directly extracted from the visual content of the frames. Moreover, the large and abrupt scale changes, the scene background changes due to the camera operation and the need of camera motion compensation make target tracking with these cameras extremely challenging. In this paper, we present a solution that provides continuous online calibration of PTZ cameras which is robust to rapid camera motion, changes of the environment due to varying illumination or moving objects. The approach also scales beyond thousands of scene landmarks extracted with the SURF keypoint detector. The method directly derives the relationship between the position of a target in the ground plane and the corresponding scale and position in the image and allows real-time tracking of multiple targets with high and stable degree of accuracy even at far distances and any zoom level.     

双目立体视觉的三维人脸重建方法

创建逼真的三维人脸模型始终是一个极具挑战性的课题.随着三维人脸模型在虚拟现实、视频监控、三维动画、人脸识别等领域的广泛应用,三维人脸重建成为计算机图像学和计算机视觉领域的一个研究热点.针对这一问题,提出一种基于双目立体视觉的三维人脸重建方法,重建过程中无需三维激光扫描仪和通用人脸模型.首先利用标定的2台摄像机获取人脸正面图像对,通过图像校正使图像对的极线对齐并且补偿摄像机镜头的畸变;在立体匹配方面,选择具有准确可靠视差的人脸边缘特征点作为种子像素,以种子像素的视差作为区域生长的视差,在外极线约束、单调性约束以及对应匹配的边缘特征点的约束下,进行水平扫描线上的区域生长,从而得到整个人脸区域的视差图,提高了对应点匹配的速度和准确度;最后,根据摄像机标定结果和立体匹配生成的视差图计算人脸空间散乱点的三维坐标,对人脸的三维点云进行三角剖分、网格细分和光顺处理.实验结果表明,该方法能够生成光滑、逼真的三维人脸模型,证明了该算法的有效性.     

灰关联分析在视频运动目标检测中的应用

由于运动图像和背景具有极大相似性,通过背景图像和运动图像之间关联程度的大小能够检测出运动目标,故提出一种基于灰关联分析的运动目标检测方法.在室内和室外不同光照场景下,通过固定摄像机捕获的视频图像序列中的运动车体和人体进行检测;选取适当的比较图像序列,对该序列和含有运动目标的视频图像作灰关联分析,以清楚、完整地提取出运动目标.该方法对背景的要求很低,对噪声的抑制能力强,可以在一定程度上抑制阴影的影响.     

立体图像对的生成

获取同一场景的立体图像对是实现双目立体成像的一个关键问题。提出了一种在三维场景已经建好的情况下生成立体图像对的方法。该方法根据双目立体视觉的原理，利用3DS MAX中的摄像机对象对场景中的物体进行坐标变换和透视投影变换，分别生成左眼视图和右眼视图。实验结果表明，两个目标摄像机与三维模型的位置关系以及基线长度是影响立体效果的重要因素，改变目标摄像机与三维模型的位置，可以分别生成正视差、负视差的立体图像对，当AB与CO的比例参数为0.05时，生成的立体图像对的立体效果较佳。