• International Journal of Contents
• /
• v.9 no.3
• /
• pp.9-14
• /
• 2013

In this paper, we will examine single face tracking algorithms with scaling function in a mobile device. Face detection and tracking either in PC or mobile device with scaling function is an unsolved problem. Standard single face tracking method with particle filter has a problem in tracking the objects where the object can move closer or farther from the camera. Therefore, we create an algorithm which can work in a mobile device and perform a scaling function. The key idea of our proposed method is to extract the average of skin color in face detection, then we compare the skin color distribution between the detected face and the tracking face. This method works well if the face position is located in front of the camera. However, this method will not work if the camera moves closer from the initial point of detection. Apart from our weakness of algorithm, we can improve the accuracy of tracking.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.6
• /
• pp.3086-3103
• /
• 2017

Video images captured by vehicle cameras often contain blurry or dithering frames due to inadvertent motion from bumps in the road or by insufficient illumination during the morning or evening, which greatly reduces the perception of objects expression and recognition from the records. Therefore, a real-time electronic stabilization method to correct fuzzy video from driving recorders has been proposed. In the first stage of feature detection, a coarse-to-fine inspection policy and a scale nonlinear diffusion filter are proposed to provide more accurate keypoints. Second, a new antiblurry binary descriptor and a feature point selection strategy for unintentional estimation are proposed, which brought more discriminative power. In addition, a new evaluation criterion for affine region detectors is presented based on the percentage interval of repeatability. The experiments show that the proposed method exhibits improvement in detecting blurry corner points. Moreover, it improves the performance of the algorithm and guarantees high processing speed at the same time.

• Journal of Magnetics
• /
• v.21 no.3
• /
• pp.322-329
• /
• 2016

In this paper, we propose an array sensing detection method for smelting of submerged arc furnaces (SAF) based on electromagnetic radiation. AC magnetic field generated by electrode currents and molten currents in the furnace is reflected outside of the furnace. According to the spatial distribution of electromagnetic field a radiation model of SAF is built. We design a 3D magnetic field sensing array system in order to collect the magnetic field information. Through the collected information, the current distribution characteristics of SAF are described and the key parameters of smelting are obtained. Theoretical simulation and field test show that the curves acquired by the sensing array can accurately reflect the information of the relative displacement when the relative displacement between the array and electrode is 10 cm. Compared with the detection method of 3D single point, the proposed array sensing method of magnetic field obtains better results in terms of real-time and accuracy, and has good practical value for industrial measurement.

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.7
• /
• pp.552-556
• /
• 2016

Obstacle detection is a key feature in the safe driving control of electric wheelchairs. The suggested obstacle detection algorithm was designed to provide obstacle avoidance direction and detect the existence of cliffs. By means of this information, the wheelchair can determine where to steer and whether to stop or go. A 3D depth camera (Microsoft KINECT) is used to scan the 3D point data of the scene, extract information on obstacles, and produce a steering direction for obstacle avoidance. To be specific, ground detection is applied to extract the obstacle candidates from the scanned data and the candidates are projected onto a 2D map. The 2D map provides discretized information of the extracted obstacles to decide on the avoidance direction (left or right) of the wheelchair. As an additional function, cliff detection is developed. By defining the "cliffband," the ratio of the predefined band area and the detected area within the band area, the cliff detection algorithm can decide if a cliff is in front of the wheelchair. Vehicle tests were carried out by applying the algorithm to the electric wheelchair. Additionally, detailed functions of obstacle detection, such as providing avoidance direction and detecting the existence of cliffs, were demonstrated.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.3
• /
• pp.248-253
• /
• 2008

This paper presents a sensing system for smart home which can detect an location transition events such as entrance/exit of a member and identify the user in a group at the same time. The proposed system is compose of two sub-systems; a wireless sensor network system and a database server system. The wireless sensing system is designed as a star network where each of sensing modules with ultrasonic sensors and a Bluetooth RF module connect to a central receiver called Bluetooth access point. We propose a method to discriminate a user by measuring the height of the user. The differences in the height of users is a key feature for discrimination. At the same time, the each sensing module can recognize whether the user goes into or out a room by using two ultrasonic sensors. The server subsystem is a sort of data logging system which read the detected event from the access point and then write it into a database system. The database system could provide the location transition information to wide range of context-aware applications for smart home easily and conveniently. We evaluate the developed method with experiments for three subjects in a family with the installation of the developed system into a real house.

• Journal of Korean Institute of Industrial Engineers
• /
• v.36 no.3
• /
• pp.154-163
• /
• 2010

Process monitoring of output variables affecting final performance have been mainly executed in semiconductor manufacturing process. However, even earlier detection of causes of output variation cannot completely prevent yield loss because a number of wafers after detecting them must be re-processed or cast away. Semiconductor manufacturers have put more attention toward monitoring process inputs to prevent yield loss by early detecting change-point of the process. In the paper, we propose the method to efficiently monitor functional input variables in multi-phase semiconductor manufacturing process. Measured input variables in the multi-phase process tend to be of functional structured form. After data pre-processing for these functional input data, change-point analysis is practiced to the pre-processed data set. If process variation occurs, key variables affecting process variation are selected using contribution plot for monitoring efficiency. To evaluate the propriety of proposed monitoring method, we used real data set in semiconductor manufacturing process. The experiment shows that the proposed method has better performance than previous output monitoring method in terms of fault detection and process monitoring.

• Journal of Information Processing Systems
• /
• v.18 no.3
• /
• pp.428-442
• /
• 2022

A reliable automatic passenger counting (APC) system is a key point in transportation related to the efficient scheduling and management of transport routes. In this study, we introduce a lightweight head detection network using deep learning applicable to an embedded system. Currently, object detection algorithms using deep learning have been found to be successful. However, these algorithms essentially need a graphics processing unit (GPU) to make them performable in real-time. So, we modify a Tiny-YOLOv3 network using certain techniques to speed up the proposed network and to make it more accurate in a non-GPU environment. Finally, we introduce an APC system, which is performable in real-time on embedded systems, using the proposed head detection algorithm. We implement and test the proposed APC system on a Samsung ARTIK 710 board. The experimental results on three public head datasets reflect the detection accuracy and efficiency of the proposed head detection network against Tiny-YOLOv3. Moreover, to test the proposed APC system, we measured the accuracy and recognition speed by repeating 50 instances of entering and 50 instances of exiting. These experimental results showed 99% accuracy and a 0.041-second recognition speed despite the fact that only the CPU was used.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.6
• /
• pp.577-582
• /
• 2015

This paper proposes an efficient detection of absolute position of a robot joint using two incremental encoders. We considers a robot joint comprising a motor, a reducer, two encoders, and a motor drive. An incremental(first) encoder provides motor's rotor position or input position of reducer while another incremental(second) encoder does output position of the reducer. A table is made where the relationship between the first and the second encoder counts is recorded. The key point is placed where the table is constructed: when a pulse occurs in the second encoder, there exists a corresponding unique count value of the first encoder. The absolute position is detected using the table by searching the second encoder position corresponding to the first encoder count value when a pulse occurs in the second encoder. The proposed method needs a small rotation, as just one second encoder's pulse angle, for the initial absolute position detection.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.05a
• /
• pp.232-235
• /
• 2018

Convolutional Neural Network is arguably the most popular deep learning architecture that is one of the most attractive area of research since it has various applications including face detection and recognition. The cascaded CNN operates at multiple resolution and rejects the background regions in the fast low resolution stages. By considering that advantage, we carry out the study on accuracy of cascaded CNN for face detection applications. The key point for our study is to analysing and improving the accuracy of cascaded CNN by applying simulations of algorithm where by we used Google's Tensorflow GPU as deep learning framework.

• Journal of Korea Water Resources Association
• /
• v.47 no.8
• /
• pp.671-684
• /
• 2014

Change point analysis is a efficient tool to understand the fundamental information in hydro-meteorological data such as rainfall, discharge, temperature etc. Especially, this fundamental information to change points to future rainfall data identified by reasonable detection skills can affect the prediction of flood and drought occurrence because well detected change points provide a key to resolve the non-stationary or inhomogeneous problem by climate change. Therefore, in this study, the comparative study to assess the performance of the 3 change point detection skills, cumulative sum (CUSUM) method, Bayesian change point (BCP) method, and segmentation by dynamic programming (DP) was performed. After assessment of the performance of the proposed detection skills using the 3 types of the synthetic series, the 2 reasonable detection skills were applied to the observed and future rainfall data at the 5 rainfall gauges in South Korea. Finally, it was suggested that BCP (with 0.9 posterior probability) could be best detection skill and DP could be reasonably recommended through the comparative study. Also it was suggested that BCP (with 0.9 posterior probability) and DP detection skills to find some change points could be reasonable at the North-eastern part in South Korea. In future, the results in this study can be efficiently used to resolve the non-stationary problems in hydrological modeling considering inhomogeneity or nonstationarity.