Fusion of Tracking Techniques to Enhance Adaptive Real-time Tracking of Arbitrary Objects
نویسندگان
چکیده
• Compute particle weights from linear SVM score of extended HOG features [2] • Baseline: Features computed for each particle • Enhancement: Sliding window detector on feature pyramid for fast re-detection after occlusions, failure detection, and to find the best negative training examples in each frame • Particles get SVM score from detector response • Sliding-window-based measurement model is less accurate (resolution is at HOG cell instead of pixel level), yellow particle in figure above will be weighted as if it was at the green position • Increases run-time efficiency, as convolution on all feature pyramid layers is faster than computing features and score per particle
منابع مشابه
Mathematical Analysis of Optimal Tracking Interval Management for Power Efficient Target Tracking Wireless Sensor Networks
In this paper, we study the problem of power efficient tracking interval management for distributed target tracking wireless sensor networks (WSNs). We first analyze the performance of a distributed target tracking network with one moving object, using a quantitative mathematical analysis. We show that previously proposed algorithms are efficient only for constant average velocity objects howev...
متن کاملAdaptive Fusion of Inertial Navigation System and Tracking Radar Data
Against the range-dependent accuracy of the tracking radar measurements including range, elevation and bearing angles, a new hybrid adaptive Kalman filter is proposed to enhance the performance of the radar aided strapdown inertial navigation system (INS/Radar). This filter involves the concept of residual-based adaptive estimation and adaptive fading Kalman filter and tunes dynamically the fil...
متن کاملA Novel Method for Tracking Moving Objects using Block-Based Similarity
Extracting and tracking active objects are two major issues in surveillance and monitoring applications such as nuclear reactors, mine security, and traffic controllers. In this paper, a block-based similarity algorithm is proposed in order to detect and track objects in the successive frames. We define similarity and cost functions based on the features of the blocks, leading to less computati...
متن کاملDoppler and bearing tracking using fuzzy adaptive unscented Kalman filter
The topic of Doppler and Bearing Tracking (DBT) problem is to achieve a target trajectory using the Doppler and Bearing measurements. The difficulty of DBT problem comes from the nonlinearity terms exposed in the measurement equations. Several techniques were studied to deal with this topic, such as the unscented Kalman filter. Nevertheless, the performance of the filter depends directly on the...
متن کاملPedestrians Tracking in a Camera Network
With the increase of the number of cameras installed across a video surveillance network, the ability of security staffs to attentively scan all the video feeds actually decreases. Therefore, the need for an intelligent system that operates as a tracking system is vital for security personnel to do their jobs well. Tracking people as they move through a camera network with non-overlapping field...
متن کامل