• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.12
• /
• pp.1762-1769
• /
• 2021

The final goal of this study is to develop a system that can analyze whether a wanted vehicle is a criminal vehicle from images collected from black boxes, smartphones, CCTVs, and so on. Data collection was collected using a self-developed black box. The used data in this study has used a total of 83,753 cases such as the eight vehicle types(truck, RV, passenger car, van, SUV, bus, sports car, electric vehicle) and 434 vehicle models. As a result of vehicle recognition using YOLO v5, mAP was found to be 80%. As a result of identifying the vehicle model with ReXNet using the self-developed black box, the accuracy was found to be 99%. The result was verified by surveying field police officers. These results suggest that improving the accuracy of data labeling helps to improve vehicle recognition performance.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.15 no.6
• /
• pp.78-90
• /
• 2011

This paper surveys the recent research activities regarding dynamical modeling of thermo-acoustic instabilities which are fundamental to actively control such phenomena in gas-turbine engines, rockets, and etc. For this, we introduce reduced-order modeling approaches, mainly conducted after 1990s. Particularly, we survey grey-box approaches, which determine the structure of the model based on physical rules and use system's input-output data for estimating parameters of the model. We also introduce black-box approaches using model structures without physics-based interpretation. Finally, we briefly discuss future directions and feasibilities of the research in this field.

• Analytical Science and Technology
• /
• v.36 no.2
• /
• pp.80-88
• /
• 2023

Fingerprints play a crucial role in the identification of potential suspects in criminal cases. However, determining the actual time, i.e., the time at which the fingermark was deposited, is challenging. Herein, we investigated the persistence and aging of fingerprints over time by observing the time evolution of latent fingerprints on a polystyrene box stored in a dark room. Fingerprint samples that were stored for up to two years could be detected with maximum accuracy using a black iron-oxide-based emulsion (black emulsion). To estimate the time of fingerprint deposition, fingerprint aging was studied by analyzing the lipid components of the fingerprints after their development. Cholesterol and squalene were selected as indicators of fingerprint aging, and their ratio was estimated to assess aging. In the case of fingerprint samples stored in a dark room for up to one month after deposition, the cholesterol/squalene ratio was approximately 0.01; it increased gradually to ≥ 0.1 over six months. A substantial reduction in the levels of cholesterol and squalene from the initial levels was also noted. Cholesterol and squalene were not detected after one year of storage. Thus, the extent of aging could be determined by analyzing the aging indicators for up to six months. Two cases that could cause error in the estimation of the fingerprint deposition time, namely, heating of the fingerprint sample before development and storage of the developed fingerprints in a dark room, were also investigated.

• International Journal of Automotive Technology
• /
• v.2 no.1
• /
• pp.17-23
• /
• 2001

Measurements of the initial values lead to an inverse and mathematically unprecisely formulated problem. A precise definition of an inverse problem is possible. It is to state a mathematical model of a physical process with clearly defined initial and exit values for the system behind the process. One can grasp the idea of an inverse problem by considering the tire as a copy of the objects of nature in a room with observations. Interpretation of nature is generally a result of an inverse problem. On one hand, the tire may be represented through the sensory organs and the nervous system as well as the experiences of the developer's existing apparatus of the projection of reality. On the other hand, it may be represented by a physical law or a model that can be confirmed or is to be refuted with the help of suitable measurements. During reconstruction of a measuring signal and the identification of a black box that can be assumed to be linear and causal, the tire becomes a first type Volterra integral equation of the convolution type. But measurements of the initial values are always fuzzy, the errors grow and the system behavior can no longer be forecasted. Thus, we have to deal with a chaotic system. This chaos produces fractals in a natural way. These are self-similar geometric structures. This self-similarity is clearly visible in the design.

• Journal of the Korean Society of Industry Convergence
• /
• v.18 no.4
• /
• pp.259-266
• /
• 2015

This study presents new scheme for various walking pattern of biped robot under the limitted enviroments. We show that the neural network is significantly more attractive intelligent controller design than previous traditional forms of control systems. A multilayer backpropagation neural network identification is simulated to obtain a learning control solution of biped robot. Once the neural network has learned, the other neural network control is designed for various trajectory tracking control with same learning-base. The main advantage of our scheme is that we do not require any knowledge about the system dynamic and nonlinear characteristic, and can therefore treat the robot as a black box. It is also shown that the neural network is a powerful control theory for various trajectory tracking control of biped robot with same learning-vase. That is, we do net change the control parameter for various trajectory tracking control. Simulation and experimental result show that the neural network is practically feasible and realizable for iterative learning control of biped robot.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2019.11a
• /
• pp.271-271
• /
• 2019

Estimating ship motion is difficult because it take place in complex environments.. Estimating ship motion is an important factor in ensuring the safety of ship, so accurate estimates are needed. Existing motion-related studies compare the apparent motion of the model acquired and the reference model by experimenting with the ship motion on a particular alignment, making it difficult to intuitively estimate the hull motion. This study introduces the concept of estimating the characteristics of ship motion as a transfer function through pole-zero interpretation and frequency response analysis by applying the method of transfer function of Linear-Time Invariant system. Ship motion analysis model using Linear-Time Invariant system is consist with 1) wave as input signal 2) ship motion as output signal 3) hull defined as black box. This model can be defined by numericalizing the ship motion as a transfer function and is expected to facilitate the characterization of the ship motion through pole-zero analysis and frequency response analysis.

• Journal of The Korean Association For Science Education
• /
• v.28 no.8
• /
• pp.955-963
• /
• 2008

This study was conducted to assess any change in students' views on the nature of science (NOS) after lessons through the activities of model construction and pattern identification. The instrument used to examine NOS views was the Views of Nature of Science questionnaire (VNOS). Four students' responses on VNOS before and after instruction were analyzed. The two levels of their views, novice and expert, were judged by the authors based on criteria set by several science educators. The instruction consisted of six hours of the so-called black box and cube activities developed for model construction and pattern identification, respectively. Students' views were at the novice level in definition of scientific theory, tentativeness of scientific knowledge, difference of hypotheses, theories and laws, model construction, and creativity and imagination in experiments and investigations. Students' views on NOS knowledge such as model and theory have improved for two students after instruction. The improvement seemed to be due to an explicit approach using the activities of model construction and pattern identification. The factors of changes and no-changes of views on NOS were identified and discussed in terms of improvement of the views.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.23 no.3
• /
• pp.515-530
• /
• 2013

The results of investigations into marine incidents have become an important basis in determining not only possible causes, but also the extent of negligence between the perpetrator and victim. However, marine incidents occur under special circumstances i.e. the marine environment, and this leads to difficulties in identifying causes due to problems in scene preservation, reenactment and acquisition of witnesses. Given the aforementioned characteristic of marine incidents, the International Convention for the Safety of Life at Sea (SOLAS) has adopted mandatory regulations on the carriage of Voyage Data Recorders (VDRs) and Automatic Identification Systems (AIS) for ships of a certain gross tonnage and upwards, so as to reflect recent developments in radio communication and marine technology. Adopted to provide an international standard for investigations and to promote cooperation, the Code of the International Standards and Recommended Practices for a Safety Investigation into a Marine Casualty or Marine Incident (Casualty Investigation Code) recommends member states to build capacity for analysis of VDR data. Against this backdrop, this paper presents methods for efficient investigations into the causes behind marine incidents based on data analysis of VDR, which serves as the black box of ships, as well as digital forensic techniques.

• Journal of Intelligence and Information Systems
• /
• v.28 no.1
• /
• pp.89-106
• /
• 2022

Recently, the amount of video data collected from smartphones, CCTVs, black boxes, and high-definition cameras has increased rapidly. According to the increasing video data, the requirements for analysis and utilization are increasing. Due to the lack of skilled manpower to analyze videos in many industries, machine learning and artificial intelligence are actively used to assist manpower. In this situation, the demand for various computer vision technologies such as object detection and tracking, action detection, emotion detection, and Re-ID also increased rapidly. However, the object detection and tracking technology has many difficulties that degrade performance, such as re-appearance after the object's departure from the video recording location, and occlusion. Accordingly, action and emotion detection models based on object detection and tracking models also have difficulties in extracting data for each object. In addition, deep learning architectures consist of various models suffer from performance degradation due to bottlenects and lack of optimization. In this study, we propose an video analysis system consists of YOLOv5 based DeepSORT object tracking model, SlowFast based action recognition model, Torchreid based Re-ID model, and AWS Rekognition which is emotion recognition service. Proposed model uses single-linkage hierarchical clustering based Re-ID and some processing method which maximize hardware throughput. It has higher accuracy than the performance of the re-identification model using simple metrics, near real-time processing performance, and prevents tracking failure due to object departure and re-emergence, occlusion, etc. By continuously linking the action and facial emotion detection results of each object to the same object, it is possible to efficiently analyze videos. The re-identification model extracts a feature vector from the bounding box of object image detected by the object tracking model for each frame, and applies the single-linkage hierarchical clustering from the past frame using the extracted feature vectors to identify the same object that failed to track. Through the above process, it is possible to re-track the same object that has failed to tracking in the case of re-appearance or occlusion after leaving the video location. As a result, action and facial emotion detection results of the newly recognized object due to the tracking fails can be linked to those of the object that appeared in the past. On the other hand, as a way to improve processing performance, we introduce Bounding Box Queue by Object and Feature Queue method that can reduce RAM memory requirements while maximizing GPU memory throughput. Also we introduce the IoF(Intersection over Face) algorithm that allows facial emotion recognized through AWS Rekognition to be linked with object tracking information. The academic significance of this study is that the two-stage re-identification model can have real-time performance even in a high-cost environment that performs action and facial emotion detection according to processing techniques without reducing the accuracy by using simple metrics to achieve real-time performance. The practical implication of this study is that in various industrial fields that require action and facial emotion detection but have many difficulties due to the fails in object tracking can analyze videos effectively through proposed model. Proposed model which has high accuracy of retrace and processing performance can be used in various fields such as intelligent monitoring, observation services and behavioral or psychological analysis services where the integration of tracking information and extracted metadata creates greate industrial and business value. In the future, in order to measure the object tracking performance more precisely, there is a need to conduct an experiment using the MOT Challenge dataset, which is data used by many international conferences. We will investigate the problem that the IoF algorithm cannot solve to develop an additional complementary algorithm. In addition, we plan to conduct additional research to apply this model to various fields' dataset related to intelligent video analysis.