Night time vehicle detection and tracking using traditional visible light cameras is a challenging task due to the limited visibility. The current state-of-the-art systems treat the vehicles at night time as paired vehicle headlights or taillights, with no ability to determine the contour of the vehicle or its spatial occupancy. Therefore, this paper proposes the first night time framework that combines the vehicle headlights and taillights to localize the vehicle contours. This new framework includes a novel multi-camera vehicle representation that groups and reconstructs vehicle headlights and taillights following mutual geometric distances between different vehicle components. This novel vehicle contour representation successfully removes duplicated vehicle lights and also compensates for the missing vehicle lights in the detection process. Eventually, vehicle headlight alignment and contour adjustment are used to further refine the vehicle contours. The proposed multi-camera system considers typical four-wheel vehicles, e.g., cars and SUVs, in the monitoring and might not be able to handle large trucks, e.g., 18-wheelers. The experiments are conducted on night time traffic videos under various scenarios and the proposed system attains an average of 0.896 in Multiple Object Tracking Accuracy (MOTA) and an average of 0.904 in Jaccard Coefficient (JC), which indicates 19.2% and 15.9% increases over the baseline system. The night time traffic datasets used in this paper are available at https://github.com/JustMeZXX/Intelligent-Night-Time-Traffic-Surveillance

Accurate three-dimensional displacement measurements of bridges and other structures have received significant attention in recent years. The main challenges of such measurements include the cost and the need for a scalable array of instrumentation. This paper presents a novel Hybrid Inertial Vision-Based Displacement Measurement (HIVBDM) system that can measure three-dimensional structural displacements by using a monocular charge-coupled device (CCD) camera, a stationary calibration target, and an attached tilt sensor. The HIVBDM system does not require the camera to be stationary during the measurements, while the camera movements, i.e., rotations and translations, during the measurement process are compensated by using a stationary calibration target in the field of view (FOV) of the camera. An attached tilt sensor is further used to refine the camera movement compensation, and better infers the global three-dimensional structural displacements. This HIVBDM system is evaluated on both short-term and long-term synthetic static structural displacements, which are conducted in an indoor simulated experimental environment. In the experiments, at a 9.75 m operating distance between the monitoring camera and the structure that is being monitored, the proposed HIVBDM system achieves an average of 1.440 mm Root Mean Square Error (RMSE) on the in-plane structural translations and an average of 2.904 mm RMSE on the out-of-plane structural translations.

Our task of video copy detection system aims to locate video segments that are partially copied or near-duplicated versions from an archive of reference videos. In 2010, video copy detection problem was sometimes considered as a solved problem, since previous research within this area used either small-scale or large-scale datasets (e.g. TRECVID 2009, Muscle-VCD) with pre-defined simulated videos. Therefore, the near-perfect results obtained on these datasets were somehow not convincing. As a result, in this paper, we introduce an effective real-scenario video copy detection system which aims to effectively and efficiently detect complex real video copies. Our system obtains decent results on a real-scenario large-scale video copy database (VCDB) generated in 2014, and measures the trade-off between effectiveness and efficiency. We believe our work can be regarded as the beginning for this challenging problem.

Ground motions associated with aircraft overflights can cover a significant portion of the seismic data collected by shallowly emplaced seismometers, such as new nodal and Distributed Acoustic Sensing systems. This article describes the first published framework for automated detection of aircraft on single channel and multichannel seismic data. The seismic data are converted to spectrograms in a sliding time window and classified as aircraft or nonaircraft in each window using a deep convolutional neural network trained with analyst-labeled data. A majority voting scheme is used to convert the output from the sequence of sliding time windows onto a decision time sequence for each channel and to combine the binary classifications on the decision time sequences across multiple channels. Precision, recall, and F-score are used to quantify the detection performance of the algorithm on nodal data using fourfold time-series cross validation. By applying our framework to data from the Sage Brush Flats nodal array in Southern California, we provide a benchmark performance and demonstrate the advantage of using an array of sensors.

Synthetically creating motion blur in two-dimensional (2D) images is a well-understood process and has been used in image processing for developing deblurring systems. There are no well-established techniques for synthetically generating arbitrary motion blur within three-dimensional (3D) images, such as depth maps and point clouds since their behavior is not as well understood. As a prerequisite, we have previously developed a method for generating synthetic motion blur in a plane that is parallel to the sensor detector plane. In this work, as a major extension, we generalize our previously developed framework for synthetically generating linear and radial motion blur along planes that are at arbitrary angles with respect to the sensor detector plane. Our framework accurately captures the behavior of the real motion blur that is encountered using a Time-of-Flight (ToF) sensor. This work uses a probabilistic model that predicts the location of invalid pixels that are typically present within depth maps that contain real motion blur. More specifically, the probabilistic model considers different angles of motion paths and the velocity of an object with respect to the image plane of a ToF sensor. Extensive experimental results are shown that demonstrate how our framework can be applied to synthetically create radial, linear, and combined radial-linear motion blur. We quantify the accuracy of the synthetic generation method by comparing the resulting synthetic depth map to the experimentally captured depth map with motion. Our results indicate that our framework achieves an average Boundary F1 (BF) score of 0.7192 for invalid pixels for synthetic radial motion blur, an average BF score of 0.8778 for synthetic linear motion blur, and an average BF score of 0.62 for synthetic combined radial-linear motion blur.