光流引导下机器人抓取目标物角点动态匹配

针对目标物位姿在线估计中的特征点动态匹配问题,以木块作为实验对象,提出了一种光流引导的目标物角点动态匹配方法。使用Yolov2-tiny进行目标物识别,完成目标物定位与分类。在提取区域内完成Shi-Tomasi角点检测,采用LK（Lucas-Kanade）光流对角点跟踪的方法实现角点在相邻两帧图像中的初始匹配,针对光流跟踪中特征点的偏移问题,对跟踪点为中心的11×11邻域进行角点再提纯。实验证明该算法能在视频序列中对目标物角点实时动态地匹配,且角点匹配结果比SIFT和Harris-SIFT方法准确度更高,实时性更好。     

Visual tracking and learning using speeded up robust features

A speeded up robust features (SURF) based optical flow algorithm is presented for visual tracking in real scenarios. SURF construct invariant features to correspond the blobs of interest across frames. Meanwhile, new feature-based optical flow algorithm is used to compute the warp matrix of a region centered on SURF key points. Furthermore, on-line visual learning for long-term tracking is performed using incremental object subspace method, which includes the correct update of the sample mean and appearance model. The proposed SURF based tracking and learning method contributes measurably to improving overall tracking performance. Experimental work demonstrates that the proposed strategy improves the performance of the classical optical flow algorithms in complicated real scenarios.     

尺度自适应在线鲁棒目标跟踪

针对在线boosting跟踪算法在目标外观发生大幅度变化以及遮挡时易产生“漂移”导致目标丢失问题进行了研究,提出一种尺度自适应在线鲁棒目标跟踪算法。算法基于目标灰度或彩色直方图统计特征构建权重图像,通过对权重图像的矩特征分析,可以实现对目标尺度的自适应调整,同时该算法引入半监督学习策略,很好地解决了由于在线学习导致的跟踪失败问题。实验结果表明,本文算法很好地解决了遮挡、目标外观和尺度变化时的鲁棒跟踪问题。与EM-shift,MIL和SPT三种算法相比,跟踪成功率以及鲁棒性均有所提高。     

基于特征自适应选择的鲁棒跟踪算法

为了解决真实场景下视频目标的跟踪问题,提出一种基于特征自适应选择的鲁棒跟踪算法。首先,针对在线AdaBoost算法特征池特征鲁棒性差的问题,提出一种基于颜色与金字塔梯度方向直方图特征相结合的特征池构造方式;然后,针对分类器在更新过程中容易受到错误样本影响的问题,对每帧跟踪结果增加遮挡检测环节以避免漂移现象的发生。大量的对比实验表明,在真实场景下所提出的方法具有更好的效果。     

利用视觉显著性和粒子滤波的运动目标跟踪

针对运动目标跟踪问题,提出一种利用视觉显著性和粒子滤波的目标跟踪算法.借鉴人类视觉注意机制的研究成果,根据目标的颜色、亮度和运动等特征形成目标的视觉显著性特征,与目标的颜色分布模型一起作为目标的特征表示模型,利用粒子滤波进行目标跟踪.该算法能够克服利用单一颜色特征所带来的跟踪不稳定问题,并能有效解决由于目标形变、光照变化以及目标和背景颜色分布相似而产生的跟踪困难问题,具有较强的鲁棒性.在多个视频序列中进行实验,并给出相应的实验结果和分析.实验结果表明,该算法用于实现运动目标跟踪是正确有效的.     

部件级表观模型的目标跟踪方法

因受遮挡、运动模糊、剧烈形变等因素的影响,稳定且准确的目标跟踪是当前计算机视觉研究领域重要挑战之一.首先采用中层视觉线索的超像素描述目标/背景的部件,以部件颜色直方图作为其特征,并通过聚类部件库的特征集构建初始表观模型,部件表达的局部性和灵活性使该模型能够准确描述目标/背景;然后,利用贝叶斯滤波模型计算目标框的初始状态,并提出相似物体干扰的检测和处理算法以避免跟踪漂移,得到更健壮的结果;最后,为了减弱形变、遮挡、模糊对表观模型的影响以更好地保持目标特征,提出一种基于部件库的特征补集的在线表观模型更新算法,根据部件变化实时反映目标/背景的变化情况.在多个具有跟踪挑战的视频序列上的实验结果表明(共12个视频序列):与现有跟踪方法相比,该算法跟踪结果的中心误差更小,成功帧数更多,能够更准确并稳定、有效地跟踪目标物体.     

一种自适应色彩融合的Mean-Shift跟踪算法

针对现有的Mean-Shift算法使用单纯的颜色特征不能适应光线及背景的变化,易受颜色相近物体干扰的问题,提出了自适应色彩融合方法来提高跟踪性能。对背景以极坐标的形式进行不等间隔采样,以融合后的目标直方图与背景直方图具有最小相似性为原则搜索色调与饱和度的最佳线性融合系数;考虑背景与目标的渐变,跟踪过程中在最佳融合系数的自适应调整邻域内调整融合系数;能够有效处理相似物体和颜色相近的大背景带来的干扰。视频序列跟踪结果表明,提出的方法能够实时、稳定地进行跟踪。     

基于相关滤波的长期跟踪算法

针对目标遮挡导致跟踪失败的问题,在核相关滤波器(KCF)的基础上,提出目标重检测的长期跟踪算法.将方向梯度直方图特征与亮度颜色空间(LAB)颜色信息融合,建立目标外观模型,并增加尺度估计对应目标尺度的改变.通过引入峰值比控制重检测模块的启动,提取Harris角点重新学习新的模型,对目标进行持续跟踪,并控制模型更新速率.在OTB数据集上的对比实验表明,文中算法跟踪精度较高,适用于遮挡情况下的目标长期跟踪.     

一种自适应色彩融合的Mean Shift跟踪算法*

针对现有的MeanShift算法使用单纯的颜色特征不能适应光线及背景的变化,易受颜色相近物体干扰的问题,提出了自适应色彩融合方法来提高跟踪性能。对背景以极坐标的形式进行不等间隔采样,以融合后的目标直方图与背景直方图具有最小相似性为原则搜索色调与饱和度的最佳线性融合系数;考虑背景与目标的渐变,跟踪过程中在最佳融合系数的自适应调整邻域内调整融合系数;能够有效处理相似物体和颜色相近的大背景带来的干扰。视频序列跟踪结果表明,提出的方法能够实时、稳定地进行跟踪。     

Unusual human behavior recognition using evolutionary technique

The vision for surveillance is an important task in many computer vision applications. The monitoring system concerns the tracking and recognition of people, and more generally, the understanding of human behaviors, from image sequences involving humans. Several methods for human tracking and human behavior recognition have been proposed by various researchers. But most of those do not have versatility and flexibility. In this paper, we propose an efficient and robust object tracking algorithm which use the color features, the distance features and count feature based on an evolutionary techniques to measure the observation similarity. And then we will track each person and classify their behavior properties by analyzing their trajectory pattern. We propose multi-layer perceptron based on hybrid genetic algorithm using Gaussian synapse make the recognition algorithm very efficient and robust for classify human behavior by trajectory pattern.     

Integrated detection and tracking of multiple faces using particle filtering and optical flow-based elastic matching

The design and implementation of a multiple face tracking framework that integrates face detection and face tracking is presented. Specifically, the incorporation of a novel proposal distribution and object shape model within the face tracking framework is proposed. A general solution that incorporates the most recent observation in the proposal distribution using a multiscale elastic matching-based optical flow algorithm is proposed. The proposed multiscale elastic matching-based optical flow algorithm is shown to be general and powerful in three significant ways. First, it allows for the tracking of both, rigid and elastic objects. Second, it enables robust tracking even in the face of sudden and gradual changes in illumination, scale and viewpoint. Third, it is suitable for tracking using both, fixed cameras and moving cameras. The proposed object shape model is based on a kernel-based line segment matching algorithm, which incorporates a voting scheme similar to the Radon Transform. The incorporation of the object shape model is shown to improve the computational complexity and accuracy of the face tracking algorithm and also enhance its robustness to occlusion, noise and scene clutter. Efficient techniques for particle sampling based on the Genetic Algorithm and for computation of the region-based likelihood function using the integral image are proposed. The incorporation of face detection within the face tracking algorithm is also proposed. Experimental results show that the proposed face tracking system is very robust in its ability to handle occlusion, noise, scene clutter and changes in illumination, scale and viewpoint and is also computationally efficient.     

基于自适应特征融合的压缩感知跟踪算法

针对运动目标外观或背景变化较大时,采用基于压缩感知的跟踪算法由于特征单一易导致漂移、跟踪不稳定甚至丢失目标等问题,提出了改进的基于自适应特征融合的压缩感知跟踪算法。该算法采用两种随机测量矩阵,分别投影V、H空间得到压缩后的纹理和颜色特征,利用在线计算的特征可靠性相对程度来自适应调整特征加权系数,充分利用两类特征的互补性来增强跟踪稳定性。对不同视频的测试结果表明,提出的方法在目标外观、背景环境变化时仍能准确跟踪目标,在目标大小为70像素×100像素时平均帧率为22帧/s,达到实时性。与提取单一特征的原压缩感知算法相比,改进后的方法在目标外观和背景变化时具有更强的鲁棒性。     

一种基于肤色与LBP特征融合的人脸跟踪算法

针对目前人脸跟踪方法易受光照变化和背景相近色的干扰,跟踪效果有时不佳或失效的问题,提出引入LBP(Local binary pattern)局部纹理特征,采用LBP直方图和颜色直方图相融合作为人脸特征描述的粒子滤波人脸跟踪方法.该方法在全局颜色和局部LBP纹理两个层次和特征线索上对人脸进行描述.实验结果表明,该方法较单一特征跟踪方法更具鲁棒性.此外,由于人脸目标的运动通常为非匀速运动,为了提高粒子传播的有效性和指导性,本文对人脸跟踪状态方程进行了改进.实验证明,改进后的人脸跟踪算法在各种复杂背景、旋转遮挡和人脸目标非匀速运动的情况下均能取得较好的跟踪效果.     

A parallel histogram-based particle filter for object tracking on SIMD-based smart cameras

We present a parallel implementation of a histogram-based particle filter for object tracking on smart cameras based on SIMD processors. We specifically focus on parallel computation of the particle weights and parallel construction of the feature histograms since these are the major bottlenecks in standard implementations of histogram-based particle filters. The proposed algorithm can be applied with any histogram-based feature sets—we show in detail how the parallel particle filter can employ simple color histograms as well as more complex histograms of oriented gradients (HOG). The algorithm was successfully implemented on an SIMD processor and performs robust object tracking at up to 30 frames per second—a performance difficult to achieve even on a modern desktop computer.     

基于相关滤波器的长时视觉目标跟踪方法

为解决相关滤波器（CF）在跟踪快速运动目标时存在跟踪失败的问题,提出一种结合了核相关滤波（KCF）跟踪器和基于光流法的检测器的长时核相关滤波（LKCF）跟踪算法。首先,使用跟踪器跟踪目标,并计算所得跟踪目标的峰值响应强度（PSR）;然后,通过比较峰值响应强度（PSR）与经验阈值大小判断目标是否跟踪丢失,当目标跟踪丢失时,在上一帧所得目标附近运用光流法检测运动目标,得到目标在当前帧中的粗略位置;最后,在此粗略位置处再次运用跟踪器得到精确位置。与核相关滤波（KCF）、跟踪-学习-检测（TLD）、压缩跟踪（CT）、时空上下文（STC）等4种跟踪算法进行对比实验,实验结果表明,所提算法在距离精确度和成功率上都表现最优,且分别比核相关滤波（KCF）跟踪算法高6.2个百分点和5.1个百分点,表明所提算法对跟踪快速运动目标有很强的适应能力。     

A precise approach to tracking dim-small targets using spectral fingerprint features

A precise method for accurately tracking dim-small targets, based on spectral fingerprint is proposed where traditional full color tracking seems impossible. A fingerprint model is presented to adequately extract spectral features. By creating a multidimensional feature space and extending the limited RGB information to the hyperspectral information, the improved precise tracking model based on a nonparametric kernel density estimator is built using the probability histogram of spectral features. A layered particle filter algorithm for spectral tracking is presented to avoid the object jumping abruptly. Finally, experiments are conducted that show that the tracking algorithm with spectral fingerprint features is accurate, fast, and robust. It meets the needs of dim-small target tracking adequately.     

基于粒子滤波后验概率分布的多特征融合跟踪

在光照和目标形变等外部条件变化的情况下,仅利用目标的单一特征难以鲁棒的跟踪目标。提出了一种基于粒子滤波后验概率分布的多特征融合跟踪算法,在粒子滤波跟踪框架下,用直方图模型表征目标的颜色和边缘特征,通过两种特征后验概率之间的"协作"与"学习"实现特征融合,各种场景的试验结果比较表明,新的融合跟踪算法比仅用单一特征跟踪、现有的多特征融合算法具有更好的稳定性和鲁棒性,特别是针对环境光照和目标背景变化较大的情况更具有优势。     

改进协方差矩阵的智能车视觉目标跟踪方法

智能车辆视觉目标具有非线性、噪声分布非高斯性的典型特点，现有算法难以实时估计目标的状态。针对识别物体复杂且多变，很难用完全的特征来描述待识别目标及其背景的不断变化，提出了一种用于融合颜色特征及SURF（Speed-Up Robust Features）特征的协方差矩阵来改进粒子滤波算法，从而提升视觉目标跟踪的实时性，满足智能车辆的要求。首先，对采集的图像进行预处理来获取感兴趣区域。接着，通过融合颜色特征及SURF特征构造范围感兴趣区域（Region Of Interest，ROI）的目标特征协方差矩阵，建立目标状态预测模型及状态观测模型，用于改进粒子滤波算法粒子重采样过程，实现对目标的精确跟踪。最后，将该方法与Mean-shift算法和颜色属性（CN）算法进行对比。实验结果表明，在智能车视觉跟踪过程中对光环境瞬时变化、目标物体存在短时遮挡以及目标物体姿态改变时，该算法在满足智能车辆对实时性要求的前提下，有效提升算法的精确度及鲁棒性。     

基于特征自动选择的Mean Shift跟踪方法

扩展传统的Mean Shift跟踪算法,使其能够实现特征和量阶自动选择。引入比率 对数图及互信息方差实现特征的自动选取,同时提出一种新的量化方法,能够更显著地区分目标和背景。实验结果表明,该算法在多场景下具有较好的鲁棒性并能提高跟踪精度,可以适应光线变化、背景干扰、被部分遮挡或色彩质量较差的情况。     

A self-adaptive image normalization and quaternion PCA based color image watermarking algorithm

This paper proposes a novel robust digital color image watermarking algorithm which combines color image feature point extraction, shape image normalization and QPCA (quaternion principal component algorithm) based watermarking embedding (QWEMS) and extraction (QWEXS) schemes. The feature point extraction method called Mexican Hat wavelet scale interaction is used to select the points which can survive various attacks and also be used as reference points for both watermarking embedding and extraction. The normalization shape image of the local quadrangle image of which the four corners are feature points of the original image is invariant to translation, rotation, scaling and skew, by which we can obtain the relationship between the feature images of the original image and the watermarked image which has suffered with geometrical attacks. The proposed QWEMS and QWEXS schemes which denote the color pixel as a pure quaternion and the feature image as a quaternion matrix can improve the robustness and the imperceptibility of the embedding watermarking. To simplify the eigen-decomposition procedure of the quaternion matrix, we develop a calculation approach with which the eigen-values and the corresponding eigen-vectors of the quaternion matrix can be computed. A binary watermark image is embedded in the principal component coefficients of the feature image. Simulation results demonstrate that the proposed algorithm can survive a variety of geometry attacks, i.e. translation, rotation, scaling and skew, and can also resist the attacks of many signal processing procedures, for example, moderate JPEG compression, salt and pepper noise, Gaussian filtering, median filtering, and so on.