改进的自适应灰度视频序列阴影检测方法

提出了一种自适应的阴影检测方法,去除了传统固定阈值阴影检测方法残留的阴影边缘,有效地改善了阴影检测效果。首先采用kmeans聚类、求前景灰度直方图峰值间平均值等方法得到自适应的阈值,在此基础上,计算满足阈值约束的前景像素点,将该点及其8邻域点都作为可去除的阴影点进行标记。最后,去除标记的阴影点及极小面积的前景区域。本文对已有的阴影检测算法进行了改进,加入了自适应的阈值计算方法并去除了原有算法检测后残留的阴影边缘,在对室内及室外视频序列进行的检测中都取得了较好的效果。      

融合颜色信息与深度信息的运动目标检测方法

基于颜色信息的运动目标检测易受光照、阴影等影响,基于深度信息的运动目标检测存在目标边缘噪声大,无法检测距离背景较近的目标等问题。针对上述问题,该文利用CCD相机获取的颜色信息及TOF相机获取的深度信息分别为每个像素建立颜色与深度信息的分类器,根据像素点的深度特征及前一帧的检测结果,自适应地为每个分类器的输出分配不同的权值,实现运动目标的检测。该文采集多组视频序列进行实验,实验结果表明该方法能有效解决单独利用颜色或深度信息进行运动目标检测时出现的问题。     

Moving cast shadow detection using online sub-scene shadow modeling and object inner-edges analysis

In this paper, we propose an adaptive and accurate moving cast shadow detection method employing online sub-scene shadow modeling and object inner-edges analysis for applications of static-camera video surveillance. To describe shadow appearance more accurately, the proposed method builds adaptive online shadow models for sub-scenes with different conditions of irradiance and reflectance. The online shadow models are learned by utilizing Gaussian functions to fit the significant peaks of accumulating histograms, which are calculated from Hue, Saturation and Intensity (HSI) difference of moving objects between background and foreground. Additionally, object inner-edges analysis is adopted to reject camouflages, which are misclassified foreground regions that are highly similar to shadows. Finally, the main shadow regions are expanded to recycle the misclassified shadow pixels based on local color constancy. The proposed algorithm can adaptively handle the shadow appearance changes and camouflages without prior information about illuminations and scenarios. Experimental results demonstrate that the proposed method outperforms state-of-the-art methods.     

一种改进的基于纹理和颜色的运动阴影检测

精确地消除活动阴影对运动目标的影响是智能视频监控的核心任务之一,针对当前运动阴影检测中采用的纹理信息过于粗糙、阈值选取需要人工干涉等问题,通过对NCC(归一化互相关)纹理算法进行改进,并结合亮度和归一化颜色特性,提出一种自适应的运动阴影检测方法。以混合高斯模型得到的前景像素为基础,通过阴影在亮度和归一化颜色的特性筛选出候选的阴影区域,结合改进的纹理算法进一步缩小阴影区域范围,最后利用空间后处理得到真实阴影。实验结果表明,该算法在有效降低噪声干扰的情况下能够较好区分局部纹理不明显的运动目标和阴影。     

Moving foreground object detection via robust SIFT trajectories

In this paper, we present an automatic foreground object detection method for videos captured by freely moving cameras. While we focus on extracting a single foreground object of interest throughout a video sequence, our approach does not require any training data nor the interaction by the users. Based on the SIFT correspondence across video frames, we construct robust SIFT trajectories in terms of the calculated foreground feature point probability. Our foreground feature point probability is able to determine candidate foreground feature points in each frame, without the need of user interaction such as parameter or threshold tuning. Furthermore, we propose a probabilistic consensus foreground object template (CFOT), which is directly applied to the input video for moving object detection via template matching. Our CFOT can be used to detect the foreground object in videos captured by a fast moving camera, even if the contrast between the foreground and background regions is low. Moreover, our proposed method can be generalized to foreground object detection in dynamic backgrounds, and is robust to viewpoint changes across video frames. The contribution of this paper is trifold: (1) we provide a robust decision process to detect the foreground object of interest in videos with contrast and viewpoint variations; (2) our proposed method builds longer SIFT trajectories, and this is shown to be robust and effective for object detection tasks; and (3) the construction of our CFOT is not sensitive to the initial estimation of the foreground region of interest, while its use can achieve excellent foreground object detection results on real-world video data.     

一种有效的图像阴影自动去除算法

目标跟踪与检测研究中,在检测运动前景时也会检测到运动目标投射的阴影。阴影使得运动目标发生几何变形,可能造成运动目标粘连,甚至造成检测不到目标。阴影去除后才能较真实的得到运动目标重心。本文研究一种利用图像YCbCr颜色信息去除阴影的方法。首先利用背景减的方法得到带影子的目标区域,其次进行YCbCr空间的背景减,由于影子和目标物体在YCbCr空间背景减信息有较大差别,因此可以通过阈值判断得到去影之后的精确目标区域,目标物体识别的精确性和鲁棒性将会得到提高。实验结果表明,该方法在去除阴影的同时又较好地保留了前景目标的信息,是一种有效的阴影去除方法。      

室内夜间微弱光源下运动目标检测和阴影去除

提出一种新的室内夜间微弱光源照明情况下的运动目标检测方法。首先进行背景建模, 获取稳固的背景图像, 之后对背景和当前帧图像进行图像增强处理, 提高其清晰度; 采用相对背景减法检测前景运动目标, 并对差分图像进行去噪和修补; 利用前景目标区域、阴影区域和背景区域像素亮度值存在差异的特点, 检测和去除背景差分图像中可能存在的阴影, 获得准确的运动目标。在室内夜间环境下采集视频进行试验, 结果验证了所提方法的有效性。     

Disocclusion filling for depth-based view synthesis with adaptive utilization of temporal correlations

The depth image-based rendering paves the path to success of 3-D video. However, one issue still remained in 3-D video is how to fill the disocclusion areas. To this end, Gaussian mixture model (GMM) is commonly employed to generate the background, and then to fill the holes. Nevertheless, GMM usually has poor performance for sequences with big foreground reciprocation. In this paper, we aim to enhance the synthesis performance. Firstly, we propose an expectation maximization based GMM background generation method, in which the pixel mixture distribution is derived. Secondly, we propose a refined foreground depth correlation approach, which recovers the background frame-by-frame based on depth information. Finally, we adaptively choose the background pixels from these two methods for filling. Experimental results show that the proposed method outperforms existing non-deep learning based hole filling methods by around 1.1 dB, and significantly surpasses deep learning based alternative in terms of subjective quality.     

基于HSV颜色空间检测与Gabor筛选器的阴影检测

针对视频分析中难以完全将前景(FG)和运动阴影正确分离,提出一种基于阴影HSV颜色空间特性与Gabor筛选器的阴影分割方法。首先,采用一种基于复杂背景(BG)的运动目标检测方法提取出运动目标;其次,采用基于HSV颜色空间阴影特性初步判定阴影区域;最后,设计基于感兴趣区域(ROI,region of interest)的Gabor筛选器对初步判定后的阴影区域进行筛选,从而检测出阴影。对不同光照和环境条件下的视频序列进行测试结果表明,方法效果好,阴影检测率高,可应用于智能视频监控的目标检测。     

基于统计的灰度视频自适应背景建模算法

针对灰度视频的目标检测依赖先验知识、召回率低以及单一算法无法同时兼顾静态与动态背景等问题,提出一种基于统计的背景建模算法。该算法无需先验知识,根据统计信息可以准确区分静态背景和动态背景,并采取不同的检测策略提取目标。对于静态背景,采用改进的三帧差分法自适应设置阈值,可以保证较高的召回率。对于动态背景,采用改进的概率密度估计法可以有效降低虚警率。采用所提算法对光照变化以及阴影进行处理,可以进一步提升算法的性能。在公开数据集与实际采集红外数据进行验证实验。实验结果表明,所提算法在多种场景中处理灰度视频的结果比其他传统算法好,在保证准确率的同时可以极大地提升召回率,并且有效提高目标的完整性。     

基于帧差法的运动车辆阴影去除算法

在视频交通车辆目标检测中,阴影问题是影响其检测准确性的关键问题之一.为了解决这个问题,提出了一种结合单模高斯模型和帧差法的运动目标阴影去除方法.首先通过单模高斯模型背景建模获取前景包括阴影在内的目标图像,再结合帧差法去除阴影.实验结果证明,该方法得到的车辆目标比较完整,并较好地去除了运动车辆阴影.     

Satellite imagery based adaptive background models and shadow suppression

Accurate segmentation of foreground objects in video scenes is critical for assuring reliable performance of vision systems for object tracking and situational awareness in outdoor scenes. Most existing techniques for background modeling and shadow suppression require that a number of parameters be “hand-tuned” based on environmental conditions. This paper presents two contributions to overcome such limitations. First, we develop and demonstrate a satellite imagery based approach for selecting appropriate background and shadow models. It is shown that the illumination conditions (i.e. cloud cover) of a scene can be reliably inferred from visible satellite images in the local region of the camera. The second contribution presented in the paper is introduction and evaluation of a Hybrid Cone-Cylinder Codebook (HC3) model which combines an adaptive efficient background model with HSV-color space shadow suppression into a single coherent framework. The structure of the HC3 model allows for seamless fusion of the satellite data. We are thereby able to exploit the fact that, for example, shadows are more pronounced on sunny days than cloudy days, allowing for more sensitive detection. The paper presents a set of experiments using day long sequences of videos from an operational surveillance system testbed. Results of these experimental analyses quantitatively illustrate the benefits of using satellite imagery to inform and adaptively adjust background and shadow modeling.     

基于组合特征和HSI颜色空间的阴影检测

为了更有效地检测视频序列中的阴影,提出基于组合特征和HSI颜色空间的阴影检测算法.对提取的前景,先采用扩展的不变矩和Gabor小波变换分别抽取待识别区域的全局特征和局部特征来建立组合特征向量,再通过建立的HSI空间的阴影颜色模型来准确检测出目标的阴影.实验结果表明,该算法具有良好的阴影检测效果.     

基于颜色空间转换的运动目标检测与阴影去除

针对室内视频监控中运动目标检测常出现的阴影误检,提出了一种基于颜色空间转换的自适应背景建模和阴影消除算法.在RGB空间采用自适应背景差对视频图像进行前景背景分离,并将检测出的前景目标锁定在活动轮廓矩形框内进行目标跟踪,对于误检的虚假目标(即阴影),利用其亮度等信息,在HSV空间去除.经实验验证,该算法对阴影的去除有良好的效果,能准确检测出真实目标.     

Robust moving shadow detection with hierarchical mixture of MLP experts

Detection and elimination of the shadows of moving objects in video sequences have been one of the major challenges in tracking applications. Since moving shadows cannot be removed from foreground by motion-based background subtraction methods, they lead to confusion and error in moving object tracking. In this paper, a novel classification method based on hierarchical mixture of experts learning for detecting shadows from foreground is proposed. A hierarchical mixture of MLP experts method (HMME) with semi-supervised teacher-directed learning (SSP-HMME) is used. It contains a two-level mixture of experts (ME) system. The main superiority of this method is that it is more robust than state-of-the-art methods in all types of indoor and outdoor environments. The robustness is against the number of light sources, illumination conditions, surface orientations, object sizes, etc., and it is estimated using accuracy rates. The video set has been collected from 7 different datasets. The results of experiments in outdoor and indoor environments show the validity of the method in the improvement on the accuracy of both detection and discrimination rate for moving shadows in video sequences. The results of the experiments show the accuracy rate of 89 % in average in different indoor and outdoor environmental conditions that is about 6 % better than current state-of-the-art methods.     

The Use of Vanishing Point for the Classification of Reflections From Foreground Mask in Videos

Extraction of foreground is a basic task in surveillance video analysis. In most real cases, its performance is heavily based on the efficiency of shadow detection and on the analysis of lighting conditions and reflections caused by mirrors or other reflective surfaces. This correspondence is focused on the improvement of foreground extraction in the case of planar reflective surfaces. We show that the geometric model of a scene with a planar reflective surface is reduced to the estimation of vanishing-point for the case of an auto-epipolar (skew-symmetric) fundamental matrix. The correspondences for the vanishing-point estimation are extracted from motion statistics. The knowledge of the position of the vanishing point allows us to integrate the geometric model and the motion statistics into image foreground-extraction to separate foreground from reflections, and thus to achieve better performance. The experiments confirm the accuracy of the vanishing point and the improvement of the foreground image mask by removing reflected object parts.      

Efficient method of moving shadow detection and vehicle classification

Moving shadow detection and removal from the extracted foreground regions of video frames, aim to limit the risk of misconsideration of moving shadows as a part of moving objects. This operation thus enhances the rate of accuracy in detection and classification of moving objects. With a similar reasoning, the present paper proposes an efficient method for the discrimination of moving object and moving shadow regions in a video sequence, with no human intervention. Also, it requires less computational burden and works effectively under dynamic traffic road conditions on highways (with and without marking lines), street ways (with and without marking lines). Further, we have used scale-invariant feature transform-based features for the classification of moving vehicles (with and without shadow regions), which enhances the effectiveness of the proposed method. The potentiality of the method is tested with various data sets collected from different road traffic scenarios, and its superiority is compared with the existing methods.     

基于稀疏表达残差的自然场景运动目标检测

本文提出一种基于稀疏表达残差的非参数化运动目标检测算法,在假设前景变化相对静态背景可以视为稀疏残差的基础上,采用视频前n帧初始化稀疏表达字典;利用字典对后续视频帧进行重构,提取每帧的重构残差;结合基于光照强度的全局阈值矩阵,将残差图像二值化,提取图像前景;利用前景区域和边缘点关系剔除ghost区域;采用增量PCA(Principal Component Analysis)算法和保守更新的思想对背景模型进行更新.在changedetection.net提供的shadow数据集上实验表明,采用全局更新和残差计算的方法,可以有效的解决由于自然场景光线变化导致的阴影变化,并且对自然场景中背景的小幅度抖动和相机抖动等问题也具有一定的抵抗能力.     

Content-Based Pseudoscopic View Detection

Stereoscopic images are generated from a pair of images (i.e., left and right images). In order to generate 3-D perception using the left and right images, it should be guaranteed that each image is perceived by the corresponding eye only. However, the depth perception becomes distorted when the left and the right eye views are interchanged, also known as a pseudoscopic problem. In this paper, we propose a novel method for detecting the pseudoscopic view by using disparity comparison in stereo images. Our approach originates from the idea that the disparities on a scene are categorized into three classes: zero disparity, positive disparity, and negative disparity, and that the foreground is usually located in front of the background. The proposed pseudoscopic view detection system consists of three sequential stages: 1) foreground/background segmentation, 2) feature points extraction, and 3) disparity comparison. We first segment the given image into two layers (i.e., foreground and background). Then, the feature points at each layer are extracted and matched to estimate the disparity characteristics of each layer. Finally, the existence of the pseudoscopic view can be investigated by using a disparity calibration model (DCM) presented in this paper and comparing the sign and magnitude of the average disparity of selected matching points set at each layer. Experimental results on various stereoscopic video sequences show that the proposed method is a useful and efficient approach in detecting the pseudoscopic view stereo images.     

一种基于Phong物体光照模型的阴影检测算法

针对目前运动目标检测算法中常将阴影误检为前景目标的问题,提出一种基于Phong物体光照模型的阴影检测算法。基于Phong物体光照模型,我们对场景中象素的亮度值进行分析,通过定义一个亮度相对变化量,推导出他在整个阴影区域是比较稳定的,所以在一个(5×5)的模板上用协方差来衡量这种稳定性,从而得到第一个阴影判决式。又推导出阴影区域亮度相对变化量随时间的变化保持相对稳定,设计一个滤波模板来增大目标区域的不稳定性,从而得到第二个阴影判决式。最后结合以上二个阴影判决式进行阴影检测,并对实验结果进行定性和定量的评估。与前人提出算法比较,本文提出的算法在阴影检测率和区分率等方面都得到了提高,具有较强的鲁棒性。