• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.7
• /
• pp.838-846
• /
• 2004

In the case where both a camera and a target are moving at the same time, the image background is successively changed, and the overlap with other moving objects is apt to be generated. The snake algorithms have been variously used in tracking the object, but it is difficult to be applied in the excessive overlap with other objects and the large bias between the snake and the target. To solve this problem, this paper presents an extended snake model. It includes an additional energy function which considers the temporal variation rate of the snake's area and a SSD algorithm which generates the template adaptive to the snake detected in the previous frame. The new energy function prevents the snake from over-shrinking or stretching and the SSD algorithm with adaptively changing template allows the prediction of the target's position in the next frame. The experimental results have shown that the proposed algorithm successfully tracks the target even when the target is temporarily occluded by other objects.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.05a
• /
• pp.123-125
• /
• 2015

Wireless ad hoc network which is comprised of wireless nodes that have the limited communication range is utilized to monitoring disaster area, tracing object, and tactical system. But in the case of wireless node on the ground, a network performance decrease because wireless channel is affected from obstacle or the node deployment is restricted. In this paper, we consider wireless network based on UAV(Unmanned Aerial Vehicle) which has little spatial constraint and quickly deploy a position. We implement test-bed included ground nodes and UAV, and measure throughput and PDR(Packet Delivery Ratio) according to the usage of UAV. We show that network performance is improved by relaying data on UAV.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1302-1308
• /
• 2019

This paper presents an approach for classifying image made by acquired position data from a 2D Lidar sensor with a convolutional neural network (CNN). Lidar sensor has been widely used for unmanned devices owing to advantages in term of data accuracy, robustness against geometry distortion and light variations. A CNN algorithm consists of one or more convolutional and pooling layers and has shown a satisfactory performance for image classification. In this paper, different types of CNN architectures based on training methods, Gradient Descent(GD) and Levenberg-arquardt(LM), are implemented. The LM method has two types based on the frequency of approximating Hessian matrix, one of the factors to update training parameters. Simulation results of the LM algorithms show better classification performance of the image data than that of the GD algorithm. In addition, the LM algorithm with more frequent Hessian matrix approximation shows a smaller error than the other type of LM algorithm.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.37 no.4
• /
• pp.11-21
• /
• 2000

This paper proposes a new algorithm for a mobile robot that detects and tracks a moving target, keeping it in the view range of the robot. The sensor system consists of a camera and a line light source. The camera system is placed so that the line light projected into the ground from the light source forms its image along the horizontal center line of the CCD array of the camera. The deformation of the line image from the horizontal center line contains the information on existence of a moving object, its velocity, and position. Since the proposed algorithm reduces the region of the image to be searched, it reduces the computational complexity significantly comparing to the conventional image processing algorithms. The performance of the proposed algorithm has been tested by implementing on a mobile robot.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.39 no.2
• /
• pp.235-240
• /
• 2015

Active Safety system is requested on the market of the medium and heavy duty commercial vehicle over 4.5ton beside the market of passenger car with advancement of the digital equipment proportionally. Unlike the passenger car, the mounting position of camera in case of the medium and heavy duty commercial vehicle is relatively high, it is disadvantaged conditions for lane recognition in contradiction to passenger car. In this work, we show the method of lane recognition through the Sobel edge, based on the spatial domain processing, Hough transform and color conversion correction. Also we suggest the low error method of front vehicles recognition in order to reduce the detection error through Haar-like, Adaboost, SVM and Template matching, etc., which are the object recognition methods by frontal camera vision. It is verified that the reliability over 98% on lane recognition is obtained through the vehicle test.

• Structural Engineering and Mechanics
• /
• v.74 no.5
• /
• pp.635-655
• /
• 2020

The sag effect of long stay cables is one of the key factors restricting further increase in the span of cable-stayed bridges. Based on the formerly proposed concept of long stay cables lifted by an auxiliary suspension cable in cross-strait cable-stayed bridges, corresponding static approximate calculations and analytical theory based on catenary and parabolic cable configurations are established. Taking a main span 1400 m cable-stayed bridge as the research object, three typical lifting conditions and the whole process of auxiliary cable lifting are analyzed and discussed. The results show that the sag effect is effectively reduced. The support efficiency is only improved when the cables are lifted above the original cable chord. Reduction of the horizontal component force of the cable is limited. The equivalent elastic modulus and the vertical support stiffness of the lifted cables are significantly increased with increased horizontal projection length and not sensitive to the change of the lifting point position. The scheme of lifting the cable to the chord midpoint is more economical because of the less steel required for the auxiliary suspension cable, but its effect on improving the vertical support efficiency is limited. The support efficiency is better when the cable is lifted to the cable end tangential to the original cable chord, but the lifting force and the cross-sectional area of the auxiliary suspension cable are doubled. The approximate calculation results of the lifted cables are very close to the numerical analysis results, which verifies the applicability of the approximation method proposed in this study. The results of parabolic approximation calculations are approximately equal to that of catenary cable geometry. As the parabolic approximation analysis theory of lifted cables is more convenient in mathematical processing, it is feasible to use parabolic approximation analysis theory as the analytical method for the conceptual design of lifted cables of super-long span cable-stayed bridges.

• Journal of IKEEE
• /
• v.11 no.4
• /
• pp.187-192
• /
• 2007

Since wireless sensor networks (WSN) have relatively narrow bandwidth and have limited memory space. Mosaic by inlaying images transmitted by adjacent sensors can provide wider field of view and smaller storage memory. Most WSN are used for surveillance purpose, image acquisition period should be sufficiently short, so that mosaic algorithm has to be run in real time. Proposed algorithm is derived by using the fact that position of sensor nodes are fixed and known. Transformation matrix can be calculated by using distance between sensor nodes and distance between sensor nodes and predefined object. Simulation result shows that proposed algorithm provides very short processing time whereas it preserves image quality.

• Journal of the Korean Society of Radiology
• /
• v.12 no.1
• /
• pp.65-70
• /
• 2018

Computed tomography system is widely used on various fields because section image of an object can be acquired. During several step to obtain section image, artifacts by many error factors can be added on the image. Ring artifact induced by the CT system is examined in this study. A test phantom of $512{\times}512$ size was constructed numerically, and the ring artifact was investigated by the algebraic reconstruction technique. The computer program was realized using Visual C++ under the fan beam geometry with projections of 720 and detector pixel of 1,280. The generation of ring artifact was verified by applying different detection efficiency on detector pixels. The ring intensity became large as increasing the ring value, and the ring artifacts were strongly emphasized near the center of the reconstructed image. The ring artifact may be eliminated by tracking the position of ring artifact on the reconstructed image and by calibrating the detector pixel.

• Journal of Korean Society of Transportation
• /
• v.28 no.4
• /
• pp.167-175
• /
• 2010

The ubiquitous transportation system environments make it possible to collect each vehicle's position and velocity data and to perform more sophisticated traffic flow management at the individual vehicle or platoon level through vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communication. It is necessary to develop a traffic flow management scheme to take advantage of the ubiquitous transportation system environments. This paper proposes an algorithm to advise the optimal speed for each vehicle according to the traffic flow condition. The algorithm aims to stabilize the traffic flow by advising the equilibrium speed to the vehicles speeding or crawling under freely flowing condition. And it aims to prevent or at least alleviate the shockwave propagation by advising the optimal speed that should dampen the speed drop under critical flow conditions. This paper builds a simulation testbed and performs some simulation experiments for the proposed algorithm. The proposed algorithm shows the expected results in terms of travel time reduction and congestion alleviation.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.3
• /
• pp.22-30
• /
• 2009

GPS and INS integrated systems are expected to become commonly available as a result of low cost Micro-Electro-Mechanical Sensor (MEMS) technology. However, the current performance achieved by low cost sensors is still relatively poor due to the large inertial sensor errors. This is particularly prevalent in the urban environment where there are significant periods of restricted sky view. To reduce the inertial sensor error, GPS and low cost INS are integrated using a Loosely Coupled Kalman Filter architecture which is appropriate in most applications where there is good satellite availability. In this paper, we present the GPS/INS sensor Integration using Loosely Coupled Kalman Filter approach. We also compare the simulation results of Wander Azimuth Strapdown Mechanization Scheme with the reference values generated by the ZH35C trajectory simulator that is describe mathematically either by the geometry of the path, or as the position of the object over time.