• IEIE Transactions on Smart Processing and Computing
• /
• v.5 no.4
• /
• pp.256-266
• /
• 2016

Concurrently detected and annotated abnormal events can have a significant impact on surveillance systems. By considering the specific domain of pedestrian trajectories, this paper presents two main contributions. First, as introduced in much of the work on trajectory-based anomaly detection in the literature, only information about pedestrian paths, such as direction and speed, is considered. Differing from previous work, this paper proposes a framework that deals with additional types of trajectory-based anomalies. These abnormal events take places when a person enters prohibited areas. Those restricted regions are constructed by an online learning algorithm that uses surrounding information, including detected pedestrians and background scenes. Second, a simple data-boosting technique is introduced to overcome a lack of training data; such a problem particularly challenges all previous work, owing to the significantly low frequency of abnormal events. This technique only requires normal trajectories and fundamental information about scenes to increase the amount of training data for both normal and abnormal trajectories. With the increased amount of training data, the conventional abnormal trajectory classifier is able to achieve better prediction accuracy without falling into the over-fitting problem caused by complex learning models. Finally, the proposed framework (which annotates tracks that enter prohibited areas) and a conventional abnormal trajectory detector (using the data-boosting technique) are integrated to form a united detector. Such a detector deals with different types of anomalous trajectories in a hierarchical order. The experimental results show that all proposed detectors can effectively detect anomalous trajectories in the test phase.

• Journal of Educational Research in Mathematics
• /
• v.27 no.2
• /
• pp.207-225
• /
• 2017

With a significant role of textbooks in shaping students' opportunities to learn, textbook analysis is essential to reveal these opportunities to learn the concept of area and volume. This research aims to show how the Korean textbooks pace students' learning of area and volume across grades by scrutinizing the textbooks with students' developmental sequences, called learning trajectories. Tasks about area and volume in all Korean elementary textbooks (grade 1 to 6) are coded with the specific developmental stages suggested in learning trajectories. As a result, we find considerable misalignment between the textbooks and the learning trajectories. The textbooks provide opportunities to experience developmental progressions of area and volume later than ages suggested in the learning trajectories. In addition, learning opportunities are significantly concentrated in grade 5 for area and grade 6 for volume with heavy emphases on applying formulas of area or volume. The findings from this research provides important implications concerning design of textbooks as well as improving students' opportunities in the mathematics classrooms.

• Journal of Environmental Science International
• /
• v.24 no.6
• /
• pp.793-806
• /
• 2015

The purpose of this study is to find out the air flow patterns affecting the PM10 concentration in Busan and the potential sources within each trajectory pattern. The synoptic air flow trajectories are classified into four clusters by HYSPLIT model and the potential sources of PM10 are estimated by PSCF model for each cluster from 2008 to 2012. The potential source locations of PM10 are compared with the distribution of PM10 anthropogenic emissions in east Asia developed in 2006 for the NASA INTEX-B mission. The annual mean concentrations of PM10 in Busan decreased from $51ug/m^3$ in 2008 to $43ug/m^3$ in 2012. The monthly mean concentrations of PM10 were high during a spring season, March to May and low during a summer season, August and September. The cluster2 composed of the air trajectories from the eastern China to Busan through the west sea showed the highest frequency, 44 %. The cluster1 composed of the air trajectories from the inner Mongolia region to Busan through the northeast area of China showed the second high frequency, 26 %. The cluster3 and 4 were composed of the trajectories originated in the southeast sea and the east sea of Busan respectively and showed low frequencies. The concentrations of in each cluster were $47ug/m^3$ in cluster1, $56ug/m^3$ in cluster2, $42ug/m^3$ in cluster3 and $37ug/m^3$ in cluster4. From these results, it was proved that the cluster1 and 2 composed of the trajectories originated in the east and northeast area of China were the causes of high PM10 concentrations in Busan. The results of PSCF and CWT model showed that the potential sources of the high PM10 concentrations were the areas of the around Mongolia and the eastern China having high emissions of PM10 from Beijing, Hebei to Shanghai through Shandong, Jiangsu.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.16 no.6
• /
• pp.1-15
• /
• 2017

Recently terrorism targets unspecified masses and causes massive destruction, which is so-called Super Terrorism. Many countries have tried hard to protect their citizens with various preparation and safety net. With inexpensive and advanced technologies of sensors, the surveillance systems have been paid attention, but few studies associated with the classification of the pedestrians' trajectories and the difference among themselves have attempted. Therefore, we collected individual trajectories at Samseoung Station using an analytical solution (system) of pedestrian trajectory by LiDAR sensor. Based on the collected trajectory data, a comprehensive framework of classifying the types of pedestrians' trajectories has been developed with data normalization and "trajectory association rule-based algorithm." As a result, trajectories with low similarity within the very same cluster is possibly detected.

• Journal of KIISE
• /
• v.43 no.1
• /
• pp.40-46
• /
• 2016

Digital road map generation is primarily based on artificial satellite photographing or in-site manual survey work. Therefore, these map generation procedures require a lot of time and a large budget to create and update road maps. Consequently, people have tried to develop automated map generation systems using GPS trajectory data sets obtained by public vehicles. A fundamental problem in this road generation procedure involves the extraction of representative trajectory such as main roads. Extracting a representative trajectory requires the base data set of piecewise line segments(GPS-trajectories), which have close starting and ending points. So, geometrically similar trajectories are selected for clustering before extracting one representative trajectory from among them. This paper proposes a new divide- and-conquer approach by partitioning the whole map region into regular grid sub-spaces. We then try to find similar trajectories by sweeping. Also, we applied the $Fr{\acute{e}}chet$ distance measure to compute the similarity between a pair of trajectories. We conducted experiments using a set of real GPS data with more than 500 vehicle trajectories obtained from Gangnam-gu, Seoul. The experiment shows that our grid partitioning approach is fast and stable and can be used in real applications for vehicle trajectory clustering.

• Proceedings of the KIEE Conference
• /
• 1990.11a
• /
• pp.421-424
• /
• 1990

An iterative learning control scheme for exact-tracking control and parameter estimation of uncertain robotic systems is presented. In the learning control structure, tracking and feedforward input converge globally and asymptotically as iteration increases. Since convergence of parameter errors depends only on the persistent exciting condition of system trajectories along the iteration independently of length of trajectories, it may be achieved with only system trajectories of small duration. In addition, these learning control schemes are expected to be effectively applicable to time-varying parametric systems as well as time-invariant systems, for the parameter estimation is performed at each fixed time along the iteration. Finally, no usage of acceleration signal and no in version of estimated inertia matrix in the parameter estimator makes these learning control schemes more feasible.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.40 no.1
• /
• pp.39-46
• /
• 1991

In the VSS control, the trajectories with a conventional sliding surface have a reaching phase which is an interval from the initial state to the first touching of the sliding surface. Since the sliding mode control can not be realized in a reaching phase, the trajectories may be sensitive to the disturbances and parameter variations. A simple nonlinear sliding surface is proposed to improve the robustness in a reaching phase. The position control of a PM DC servo motor using a new sliding surface is carried out and is compared to the one using the conventional surface. The sliding mode occurs in entire trajectories with the proposed new sliding surface and the improved robustness is obtained.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1995.10a
• /
• pp.315-318
• /
• 1995

In this paper, an optimal trunk trajectory for stable walking of biped robots is expressed as a simple differential equation, which is then solved by numerical methods. We used ZMP (Zero Moment Point), the virtual total ground reaction point within the region of the supporting food, as the criterion of stability of biped robot walking. If the ZMP is located outside of the stable region in dynamic walking, biped robots fall down. The biped robot considered in this paper consists of two legs and a trunk. The trajectories of the two legs and the ZMP of the biped robot are determined such that they are similar ti those of a human. Based upon those trajectories, the trunk trajectory is solved by numerically integrating differential dynamic equations. Leg motions are controlled by the computed torque control method. The effectiveness of control algorithm as well as the trajectories is confirmed by computer simulations.

• Journal of Korean Society for Atmospheric Environment
• /
• v.7 no.2
• /
• pp.89-95
• /
• 1991

Yellow sands often occur in Korea during April and May each year, and they are believed to come from the Mongolian Gobi desert as the snow starts to melt in Spring time. Since the analysis of aerosol particulates can hardly distinguish the origin of particulates, the isentropic analysis of meteorological data is often used for the trajectories of the long range transport of yellows sand or air pollutants. The yellow sand case of April 9 $\sim$ 15, 1988, in Korea is analyzed for the identification of long range transport of yellow sands and their trajectories in East Asia, using isentropic analyses. We have tranformed the ECMWF grid data, analyzed in pressure coordinates, into the isentropic coordinates and then have traced the 286 K and 290 K air mass which started Gobi desert. The result shows the transport of yellow sands from the Gobi desert to the Korean peninsula.

• Journal of Astronomy and Space Sciences
• /
• v.25 no.4
• /
• pp.361-374
• /
• 2008

Recently, satellite formation flying has been a topic of significant research interest in aerospace society because it provides potential benefits compared to a large spacecraft. Some techniques have been proposed to design optimal formation trajectories minimizing fuel consumption in the process of formation configuration or reconfiguration. In this study, a method is introduced to build fuel-optimal trajectories minimizing a cost function that combines the total fuel consumption of all satellites and assignment of fuel consumption rate for each satellite. This approach is based on collocation and nonlinear programming to solve constraints for collision avoidance and the final configuration. New constraints of nonlinear equality or inequality are derived for final configuration, and nonlinear inequality constraints are established for collision avoidance. The final configuration constraints are that three or more satellites should form a projected circular orbit and make an equilateral polygon in the horizontal plane. Example scenarios, including these constraints and the cost function, are simulated by the method to generate optimal trajectories for the formation configuration and reconfiguration of multiple satellites.