A method for moving object recognition and tracking in the intelligent traffic monitoring system is presented. For the shortcomings and deficiencies of the frame-subtraction method, a redundant discrete wavelet transf...A method for moving object recognition and tracking in the intelligent traffic monitoring system is presented. For the shortcomings and deficiencies of the frame-subtraction method, a redundant discrete wavelet transform （RDWT） based moving object recognition algorithm is put forward, which directly detects moving objects in the redundant discrete wavelet transform domain. An improved adaptive mean-shift algorithm is used to track the moving object in the follow up frames. Experimental results show that the algorithm can effectively extract the moving object, even though the object is similar to the background, and the results are better than the traditional frame-subtraction method. The object tracking is accurate without the impact of changes in the size of the object. Therefore the algorithm has a certain practical value and prospect.展开更多
Crowd behaviors analysis is the‘state of art’research topic in the field of computer vision which provides applications in video surveillance to crowd safety,event detection,security,etc.Literature presents some of ...Crowd behaviors analysis is the‘state of art’research topic in the field of computer vision which provides applications in video surveillance to crowd safety,event detection,security,etc.Literature presents some of the works related to crowd behavior detection and analysis.In crowd behavior detection,varying density of crowds and motion patterns appears to be complex occlusions for the researchers.This work presents a novel crowd behavior detection system to improve these restrictions.The proposed crowd behavior detection system is developed using hybrid tracking model and integrated features enabled neural network.The object movement and activity in the proposed crowded behavior detection system is assessed using proposed GSLM-based neural network.GSLM based neural network is developed by integrating the gravitational search algorithm with LM algorithm of the neural network to increase the learning process of the network.The performance of the proposed crowd behavior detection system is validated over five different videos and analyzed using accuracy.The experimentation results in the crowd behavior detection with a maximum accuracy of 93%which proves the efficacy of the proposed system in video surveillance with security concerns.展开更多
In this paper we discuss a kind of multitarget tracking and association method based on the data fusion of heterogeneous multiple feature data gained by a sensor such as space state, signal amplitude, Doppler frequenc...In this paper we discuss a kind of multitarget tracking and association method based on the data fusion of heterogeneous multiple feature data gained by a sensor such as space state, signal amplitude, Doppler frequency and so on. In order to introduce quantitatively those heterogeneous multiple feature data which are possibly gained by a sensor into the discussion of tracking and association problem, we define a correlation measure which we explain as the generalization of conventional association decision. In conventional Nearest Neighbor method, the decision function can take only two values, 1 or 0, to represent the decision of association or not association. In our method, correlation measure can be take any real value from 0 to 1 to represent the extent of correlation. Considering the practical circumstances that some feature data might not be easily gained continuously, we introduce an effective factor to deal with these cases. In the paper we also discuss the comparative computer simulation tests and give the results.展开更多
The JTC technology deals with the problem of target tracking and target classification simultaneously within a unified framework.The fundamental idea of the JTC technology is that by taking advantage of the mutual exc...The JTC technology deals with the problem of target tracking and target classification simultaneously within a unified framework.The fundamental idea of the JTC technology is that by taking advantage of the mutual exchange of useful information between the tracker and classifier,significant improvements in performance of both target tracking and target classification can be expected.The principle of JTC technology is introduced.The existing JTC technologies are broadly categorized into two classes,i.e.,point-target-motion-model-based JTC and rigid-target-motion-based JTC,which are then compared in detail.The advance of the JTC technology is surveyed with comments on some related literatures.Finally,some opening topics of the JTC technology are discussed.展开更多
目标跟踪作为图像处理领域的重要组成部分,广泛应用于智能视频监控、军事侦察等领域。但在面对物体形变以及遮挡等复杂应用场景时,相关滤波算法由于缺乏目标和背景判别区分以及遮挡状态判断等策略,存在跟错目标、缓慢漂移到背景等现象,...目标跟踪作为图像处理领域的重要组成部分,广泛应用于智能视频监控、军事侦察等领域。但在面对物体形变以及遮挡等复杂应用场景时,相关滤波算法由于缺乏目标和背景判别区分以及遮挡状态判断等策略,存在跟错目标、缓慢漂移到背景等现象,在遮挡后目标重新出现时,缺乏重检测机制,这些问题导致了跟踪性能在实际工程中大幅下降。针对以上问题进行改进设计,首先在跟踪过程中,使用网络优化器更新多层深度特征提取网络,优化损失函数提高目标与背景的判别能力;其次,采用多重检测抗遮挡优化机制,确定跟踪器状态更新机制;最后,基于深度学习进行检测跟踪识别一体化设计,实现跟踪前典型目标的自动捕获,目标受遮挡后重新出现时实现对典型目标的重新捕获定位。在实验分析中,分别从跟踪精度、可视化定量损失以及算法速度等方面进行了性能验证。实测数据显示,本文采用的方法在以上方面性能表现良好,优于改进前的ECO(efficientconvolution operators for tracking)算法。展开更多
文摘A method for moving object recognition and tracking in the intelligent traffic monitoring system is presented. For the shortcomings and deficiencies of the frame-subtraction method, a redundant discrete wavelet transform （RDWT） based moving object recognition algorithm is put forward, which directly detects moving objects in the redundant discrete wavelet transform domain. An improved adaptive mean-shift algorithm is used to track the moving object in the follow up frames. Experimental results show that the algorithm can effectively extract the moving object, even though the object is similar to the background, and the results are better than the traditional frame-subtraction method. The object tracking is accurate without the impact of changes in the size of the object. Therefore the algorithm has a certain practical value and prospect.
文摘Crowd behaviors analysis is the‘state of art’research topic in the field of computer vision which provides applications in video surveillance to crowd safety,event detection,security,etc.Literature presents some of the works related to crowd behavior detection and analysis.In crowd behavior detection,varying density of crowds and motion patterns appears to be complex occlusions for the researchers.This work presents a novel crowd behavior detection system to improve these restrictions.The proposed crowd behavior detection system is developed using hybrid tracking model and integrated features enabled neural network.The object movement and activity in the proposed crowded behavior detection system is assessed using proposed GSLM-based neural network.GSLM based neural network is developed by integrating the gravitational search algorithm with LM algorithm of the neural network to increase the learning process of the network.The performance of the proposed crowd behavior detection system is validated over five different videos and analyzed using accuracy.The experimentation results in the crowd behavior detection with a maximum accuracy of 93%which proves the efficacy of the proposed system in video surveillance with security concerns.
文摘In this paper we discuss a kind of multitarget tracking and association method based on the data fusion of heterogeneous multiple feature data gained by a sensor such as space state, signal amplitude, Doppler frequency and so on. In order to introduce quantitatively those heterogeneous multiple feature data which are possibly gained by a sensor into the discussion of tracking and association problem, we define a correlation measure which we explain as the generalization of conventional association decision. In conventional Nearest Neighbor method, the decision function can take only two values, 1 or 0, to represent the decision of association or not association. In our method, correlation measure can be take any real value from 0 to 1 to represent the extent of correlation. Considering the practical circumstances that some feature data might not be easily gained continuously, we introduce an effective factor to deal with these cases. In the paper we also discuss the comparative computer simulation tests and give the results.
文摘The JTC technology deals with the problem of target tracking and target classification simultaneously within a unified framework.The fundamental idea of the JTC technology is that by taking advantage of the mutual exchange of useful information between the tracker and classifier,significant improvements in performance of both target tracking and target classification can be expected.The principle of JTC technology is introduced.The existing JTC technologies are broadly categorized into two classes,i.e.,point-target-motion-model-based JTC and rigid-target-motion-based JTC,which are then compared in detail.The advance of the JTC technology is surveyed with comments on some related literatures.Finally,some opening topics of the JTC technology are discussed.
文摘目标跟踪作为图像处理领域的重要组成部分,广泛应用于智能视频监控、军事侦察等领域。但在面对物体形变以及遮挡等复杂应用场景时,相关滤波算法由于缺乏目标和背景判别区分以及遮挡状态判断等策略,存在跟错目标、缓慢漂移到背景等现象,在遮挡后目标重新出现时,缺乏重检测机制,这些问题导致了跟踪性能在实际工程中大幅下降。针对以上问题进行改进设计,首先在跟踪过程中,使用网络优化器更新多层深度特征提取网络,优化损失函数提高目标与背景的判别能力;其次,采用多重检测抗遮挡优化机制,确定跟踪器状态更新机制;最后,基于深度学习进行检测跟踪识别一体化设计,实现跟踪前典型目标的自动捕获,目标受遮挡后重新出现时实现对典型目标的重新捕获定位。在实验分析中,分别从跟踪精度、可视化定量损失以及算法速度等方面进行了性能验证。实测数据显示,本文采用的方法在以上方面性能表现良好,优于改进前的ECO(efficientconvolution operators for tracking)算法。
文摘近年来,随着羊只养殖向大规模和精细化的方向发展,羊场对智能化管理的需求日益增加。因此,精准的个体识别和行为监测变得尤为重要,对多目标跟踪(Multiple object tracking, MOT)算法的准确性提出了更高要求。然而,现有的MOT算法在目标遮挡和动态场景下的性能仍不理想。本文提出两种跟踪线索:深度调制交并比(Depth modulated intersection over union, DIoU)和轨迹方向建模(Tracklet direction modeling, TDM),旨在补充交并比(Intersection over union, IoU)线索,提高多目标跟踪的精准度和鲁棒性。DIoU线索通过引入目标的深度信息改进了传统的IoU计算方法。TDM聚焦于目标的运动趋势,预测其未来的移动方向。本文将DIoU和TDM跟踪线索集成到BoT-SORT算法中,形成改进的多目标跟踪算法。在两个私有数据集上,改进算法相比基线方法,MOTA(Multiple object tracking accuracy)指标分别提高1.6、1.7个百分点,IDF1(Identification F1 score)指标分别提高1.9、1.0个百分点。结果显示,改进算法在复杂场景中的跟踪连续性和准确性显著提升。
