• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.4
• /
• pp.34-43
• /
• 2010

Texture information plays an important role in object recognition and classification. To perform an accurate classification, the texture feature used in the classification must be highly discriminative. This paper presents a novel texture descriptor for texture-based image retrieval and its application in Computer-Aided Diagnosis (CAD) system for Emphysema classification. The texture descriptor is based on the combination of local surrounding neighborhood difference and centralized neighborhood difference and is named as Combined Neighborhood Difference (CND). The local differences of surrounding neighborhood difference and centralized neighborhood difference between pixels are compared and converted into binary codewords. Then binomial factor is assigned to the codewords in order to convert them into high discriminative unique values. The distribution of these unique values is computed and used as the texture feature vectors. The texture classification accuracies using Outex and Brodatz dataset show that CND achieves an average of 92.5%, whereas LBP, LND and Gabor filter achieve 89.3%, 90.7% and 83.6%, respectively. The implementations of CND in the computer-aided diagnosis of Emphysema is also presented in this paper.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.1
• /
• pp.38-48
• /
• 1996

We implemented and evaluated the preformance of a task level pipelined multicomputer "RV860-PIPE(Realtime Vision i860 system using PIPEline)" for computer vision applications. RV860-PIPE is a message-passing MIMD computer having ring interconnection network which is appropriate for vision processing. We designed the node computer of RV860-PIPE using a 64-bit microprocessor to have generality and high processing power for various vision algorithms. Furthermore, to reduce the communication overhead between node computers and between node computer and a frame grabber, we designed dedicated high speed communication channels between them. We showed the practical applicability of the implemented system by evaluting performances of various computer vision applications like edge detection, real-time moving object tracking, and real-time face recognition.

• Journal of Digital Convergence
• /
• v.16 no.12
• /
• pp.441-447
• /
• 2018

This study aims to develop digital techniques that enable repeated practice of dental radiography using augmented reality technology. A three-dimensional object was fabricated by superimposing a photograph of an adult model and a computed tomography image of a manikin phantom. The system was structured using 106 radiographs such that one of these saved radiographs is opened when the user attempts to take a radiograph on a mobile device. This system enabled users to repeatedly practice at the pre-clinical stage without exposure to radiation. We attempt to contribute to enhancing dental hygienists' competency in dental radiography using these techniques. However, a system that enables the user to actually take a radiograph based on face recognition would be more useful in terms of practice, so additional studies are needed on the topic.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.1
• /
• pp.44-49
• /
• 2021

Recently, due to COVID-19, studies have been popularly worked to apply neural network to mask wearing automatic detection system. For applying neural networks, the 1-stage detection or 2-stage detection methods are used, and if data are not sufficiently collected, the pretrained neural network models are studied by applying fine-tuning techniques. In this paper, the system is consisted of 2-stage detection method that contain MTCNN model for face recognition and ResNet model for mask detection. The mask detector was experimented by applying five ResNet models to improve accuracy and fps in various environments. Training data used 17,217 images that collected using web crawler, and for inference, we used 1,913 images and two one-minute videos respectively. The experiment showed a high accuracy of 96.39% for images and 92.98% for video, and the speed of inference for video was 10.78fps.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.5
• /
• pp.17-22
• /
• 2022

In this paper, we design a lightweight embedded device that can support intelligent edge computing, and show that the device quickly detects an object in an image input from a camera device in real time. The proposed system can be applied to environments without pre-installed infrastructure, such as an intelligent video control system for industrial sites or military areas, or video security systems mounted on autonomous vehicles such as drones. The On-Device AI(Artificial intelligence) technology is increasingly required for the widespread application of intelligent vision recognition systems. Computing offloading from an image data acquisition device to a nearby edge device enables fast service with less network and system resources than AI services performed in the cloud. In addition, it is expected to be safely applied to various industries as it can reduce the attack surface vulnerable to various hacking attacks and minimize the disclosure of sensitive data.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.19 no.2
• /
• pp.69-78
• /
• 2023

In this paper, IO+5G dedicated hardware is developed and an AI device communication system equipped with a 5G is designed and tested. The AI device communication system equipped with a 5G receives the collected real-time images and the information collected from the IoT sensor in real time is to analyze the information and generates the risk detection events in the AI processing board. The event generated in the AI processing board creates a 5G channel in the dedicated hardware equipped with IO+5G. The created 5G channel delivers event video to the control video server. The 5G based dongle network enables faster data collection and more precise data measurement compared to wireless LAN and 5G routers. As a result of the experiment in this paper, the average test result of the 5G dongle network is about 51% faster than the Wi-Fi average test result in downlink and about 40% faster in uplink. In addition, when comparing the test result with terms of the 5G rounter to be set to 80% upload and 20% download, the average test result is that the 5G dongle network is about 11.27% faster when downloading and about 17.93% faster when uploading. when comparing the test result with terms of the the router to be set to 60% upload and 40% download, the 5G dongle network is about 11.19% faster when downlinking and about 13.61% faster when uplinking. Therefore, in this paper it describes that the developed 5G dongle network can improve the results by collecting data and analyzing it faster than wireless LAN and 5G routers.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.05a
• /
• pp.309-311
• /
• 2014

Despite increasing demands for big data application based on the analysis of scattered unstructured data, few relevant studies have been reported. Accordingly, the present study suggests a technique enabling a sentence-based semantic analysis by extracting objects from collected web information and automatically analyzing the relationships between such objects with collective intelligence and language processing technology. To be specific, collected information is stored in DBMS in a structured form, and then morpheme and feature information is analyzed. Obtained morphemes are classified into objects of interest, marginal objects and objects of non-interest. Then, with an inter-object attribute recognition technique, the relationships between objects are analyzed in terms of the degree, scope and nature of such relationships. As a result, the analysis of relevance between the information was based on certain keywords and used an inter-object relationship extraction technique that can determine positivity and negativity. Also, the present study suggested a method to design a system fit for real-time large-capacity processing and applicable to high value-added services.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.101-125
• /
• 2022

Recently, with the development of computing technology and the improvement of the cloud environment, deep learning technology has developed, and attempts to apply deep learning to various fields are increasing. A typical example is anomaly detection, which is a technique for identifying values or patterns that deviate from normal data. Among the representative types of anomaly detection, it is very difficult to detect a contextual anomaly that requires understanding of the overall situation. In general, detection of anomalies in image data is performed using a pre-trained model trained on large data. However, since this pre-trained model was created by focusing on object classification of images, there is a limit to be applied to anomaly detection that needs to understand complex situations created by various objects. Therefore, in this study, we newly propose a two-step pre-trained model for detecting abnormal situation. Our methodology performs additional learning from image captioning to understand not only mere objects but also the complicated situation created by them. Specifically, the proposed methodology transfers knowledge of the pre-trained model that has learned object classification with ImageNet data to the image captioning model, and uses the caption that describes the situation represented by the image. Afterwards, the weight obtained by learning the situational characteristics through images and captions is extracted and fine-tuning is performed to generate an anomaly detection model. To evaluate the performance of the proposed methodology, an anomaly detection experiment was performed on 400 situational images and the experimental results showed that the proposed methodology was superior in terms of anomaly detection accuracy and F1-score compared to the existing traditional pre-trained model.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.8A
• /
• pp.639-647
• /
• 2012

Vision-based on-road vehicle detection is one of the key techniques in automotive driver assistance systems. However, due to the huge within-class variability in vehicle appearance and environmental changes, it remains a challenging task to develop an accurate and reliable detection system. In general, a vehicle detection system consists of two steps. The candidate locations of vehicles are found in the Hypothesis Generation (HG) step, and the detected locations in the HG step are verified in the Hypothesis Verification (HV) step. Since the final decision is made in the HV step, the HV step is crucial for accurate detection. In this paper, we propose using a reduced multivariate polynomial pattern classifier (RM) for the HV step. Our experimental results show that the RM classifier outperforms the well-known Support Vector Machine (SVM) classifier, particularly in terms of the fast decision speed, which is suitable for real-time implementation.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.5
• /
• pp.1070-1075
• /
• 2013

In this research, we describe a security robot system and ongoing research results to control human's wrong direction in order to forbid human to enter security zone. Proposed robot system surveils a security area with equipped laserscanner sensor usually. When it detect walking human who is for the area, robot calculates his velocity vector, plans own path to forestall and interrupts him who want to head restricted area and starts to move along the estimated trajectory. The walking human is assumed to be a point-object and projected onto an scanning plane to form a geometrical constraint equation that provides position data of the human based on the kinematics of the mobile robot. While moving the robot continues these processes for adapting change of situation. After arriving at an opposite position human's walking direction, the robot advises him not to be headed more and change his course. The experimental results of estimating and tracking of the human in the wrong direction with the mobile robot are presented.