• 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 Korean Society for Geospatial Information Science
• /
• v.24 no.1
• /
• pp.69-80
• /
• 2016

Some studies were conducted for the purpose of minimizing total clearance time for rapid evacuation from the indoor spaces when disaster occurs. Most studies took a long time to calculate the optimal evacuation route that derived minimum evacuation time. For this reason, this study proposes an evacuation route assignment algorithm that can shorten the total clearance time in a short operational time. When lots of exits are in the building, this algorithm can shorten the total clearance time by assigning the appropriate pedestrian traffic volume to each exit and balances each exit-load. The graph theory and greedy algorithm were utilized to assign pedestrian traffic volume to each exit in this study. To verify this algorithm, study used a cellular automata-based evacuation simulator and experimented various occupants distribution in a building structure. As a result, the total clearance time is reduced by using this algorithm, compared to the case of evacuating occupants to the exit within shortest distance. And it was confirmed that the operation takes a short time In a large building structure.

• Science of Emotion and Sensibility
• /
• v.13 no.1
• /
• pp.197-206
• /
• 2010

For an emotion retrieval system implementation to support pedestrian navigation, coordinating the pedestrian emotion model with the system user's emotion is considered a key component. This study proposes a new method for capturing the user's model that corresponds to the pedestrian emotion model and examines the validity of the method. In the first phase, a database comprising a set of interior images that represent hypothetical destinations was developed. In the second phase, 10 subjects were recruited and asked to evaluate on navigation and satisfaction toward each interior image in five rounds of navigation experiments. In the last phase, the subjects' feedback data was used for of the pedestrian emotion model, which is called ‘learning' in this study. After evaluations by the subjects, the learning effect was analyzed by the following aspects: recall ratio, precision ratio, retrieval ranking, and satisfaction. Findings of the analysis verify that all four aspects significantly were improved after the learning. This study demonstrates the effectiveness of the learning algorithm for the proposed pedestrian emotion model. Furthermore, this study demonstrates the potential of such pedestrian emotion model to be well applicable in the development of various mobile contents service systems dealing with visual images such as commercial interiors in the future.

• Journal of the HCI Society of Korea
• /
• v.14 no.2
• /
• pp.41-47
• /
• 2019

In this study, a smart assistive device is designed to recognize pedestrian signal and to provide audio instructions for visually impaired people in crossing streets safely. Walking alone is one of the biggest challenges to the visually impaired and it deteriorates their life quality. The proposed device has a camera attached on a pair of glasses which can detect traffic lights, recognize pedestrian signals in real-time using a machine learning algorithm on GPU board and provide audio instructions to the user. For the portability, the dimension of the device is designed to be compact and light but with sufficient battery life. The embedded processor of device is wired to the small camera which is attached on a pair of glasses. Also, on inner part of the leg of the glasses, a bone-conduction speaker is installed which can give audio instructions without blocking external sounds for safety reason. The performance of the proposed device was validated with experiments and it showed 87.0% recall and 100% precision for detecting pedestrian green light, and 94.4% recall and 97.1% precision for detecting pedestrian red light.

• Journal of Computing Science and Engineering
• /
• v.8 no.4
• /
• pp.207-214
• /
• 2014

In pedestrian detection applications, one of the most popular frameworks that has received extensive attention in recent years is widely known as a 'Hough forest' (HF). To improve the accuracy of detection, this paper proposes a novel split function to exploit the statistical information of the training set stored in each node during the construction of the forest. The proposed split function makes the trees in the forest more robust to noise and illumination changes. Moreover, the errors of each stage in the training forest are minimized using a global loss function to support trees to track harder training samples. After having the forest trained, the standard HF detector follows up to search for and localize instances in the image. Experimental results showed that the detection performance of the proposed framework was improved significantly with respect to the standard HF and alternating decision forest (ADF) in some public datasets.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.8 no.5
• /
• pp.968-972
• /
• 2007

The importance of bumper system lies not only in the styling of vehicles, but also in the protection of vehicles and pedestrians from reasonable impact. In this study, we proceed to search a method for efficient bumper system without using the impact test and the computer simulation to analyze the bumper system. In the process of the research, we proposed the each method that is used to search the shape that satisfy each regulations in first, because bumper's dimensions to satisfy 'vehicle protection' regulations and form's dimensions to satisfy ‘pedestrian protection' regulations are difficult to exist together. After that we proposed the calculate method and design algorithm that is used to search a reasonable point satisfying the two regulations together.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.3
• /
• pp.840-860
• /
• 2023

Aiming at the serious occlusion and slow tracking speed in pedestrian target tracking and recognition in complex scenes, a target tracking method based on improved YOLO v5 combined with Deep SORT is proposed. By merging the attention mechanism ECA-Net with the Neck part of the YOLO v5 network, using the CIoU loss function and the method of CIoU non-maximum value suppression, connecting the Deep SORT model using Shuffle Net V2 as the appearance feature extraction network to achieve lightweight and fast speed tracking and the purpose of improving tracking under occlusion. A large number of experiments show that the improved YOLO v5 increases the average precision by 1.3% compared with other algorithms. The improved tracking model, MOTA reaches 54.3% on the MOT17 pedestrian tracking data, and the tracking accuracy is 3.7% higher than the related algorithms and The model presented in this paper improves the FPS by nearly 5 on the fps indicator.

• Journal of Digital Convergence
• /
• v.11 no.4
• /
• pp.243-249
• /
• 2013

In this paper, GPS(Global Positioning System) that can be used outdoors and GPS(Global Positioning System) is not available for indoor Wi-Fi(Wireless Fidelity) using the Android-based location information system has been designed and implemented. Pedestrians in a room in order to estimate the location of the pedestrian's position, regardless of need to obtain the absolute position and relative position, depending on the movement of pedestrians in a row it is necessary to estimate. In order to estimate the initial position of the pedestrian Wi-Fi Fingerprinting was used. Most existing Wi-Fi Fingerprinting position error small WKNN(Weighted K Nearest Neighbor) algorithm shortcoming EWKNN (Enhanced Weighted K Nearest Neighbor) using the algorithm raised the accuracy of the position. And in order to estimate the relative position of the pedestrian, the smart phone is mounted on the IMUInertial Measurement Unit) because the use did not require additional equipment.

• Journal of Korean Society of Transportation
• /
• v.22 no.7 s.78
• /
• pp.147-154
• /
• 2004

The purpose of this paper is to improve the existing phase split algorithm considering the minimum green time in COSMOS. In the case of a signalized intersection where two wide and narrow streets intersect each other, the time required for the pedestrian crossing is frequently longer than the time alloted to the through traffic on a minor street. In order to meet the minimum green time requirement for the pedestrian less time in alloted automatically to the left-turn traffic, creating heavy congestion on the left-turn approach. To solve this problem, this study suggests a new algorithm which shares the barrier using minimum green time and shares the burden with signal phases alloted to the crossing street traffic on the basis of the equal ratio of the degree of saturation, while maintaining the minimum green time requirement. The new algorithm was compared with the existing algorithm by using a microscopic simulation model for COSMOS evaluation developed at Ajou University. The simulation results show that the new algorithm produces better performance than the existing one.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.7
• /
• pp.640-645
• /
• 2013

This paper presents a step length estimation algorithm for Pedestrian Dead Reckoning using linear calibrated ZUPT (zero velocity update) with a foot mounted IMU. The IMU consists of 3 axis accelerometer, gyro and magnetometer. Attitude of IMU is estimated using an inertial navigation algorithm. To increase accuracy of step length estimation algorithm, we propose a stance detection algorithm and an enhanced ZUPT. The enhanced ZUPT calculates firefighter's step length considering velocity error caused by sensor bias during one step. This algorithm also works efficiently at various motions, such as crawling, sideways and stair stepping. Through experiments, the step length estimation performance of the proposed algorithm is verified.