• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.94-98
• /
• 2009

A video caption recognition system has been developed for broadcasting sport events such as major league baseball. The purpose of the system is to translate the information expressed in English units such as miles per hour (MPH) to the international system of units (SI) such as km/h. The system detects the ball speed displayed in the video and recognizes the numerals. The ball speed is then converted to km/h and displayed by the following character generator (CG) system. Although neural-network based methods are widely used for character and numeral recognition, we use template matching to avoid the training process required before the broadcasting. With the proposed template matching method, the operator can cope with the situation when the caption’s appearance changed without any notification. Templates are configured by the operator with a captured screenshot of the first pitch with ball speed. Templates are updated with following correct recognition results. The accuracy of the recognition module is over 97%, which is still not enough for live broadcasting. When the recognition confidence is low, the system asks the operator for the correct recognition result. The operator chooses the right one using hot keys.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.1
• /
• pp.71-77
• /
• 2014

In this paper, we propose an improved numeric character recognition method which can recognize numeric characters well under low-illuminated and shade-illuminated environment. The LN(Local Normalization) preprocessing method is used in order to enhance low-illuminated and shade-illuminated image quality. The reading area is detected using line segment information extracted from the illumination-normalized meter images, and then the three-phase procedures are performed for segmentation of numeric characters in the reading area. Finally, an efficient hybrid classifier is used to classify the segmented numeric characters. The proposed numeric character classifier is a combination of multi-layered feedforward neural network and template matching module. Robust heuristic rules are applied to classify the numeric characters. Experiments using meter image database were conducted. Meter image database was made using various kinds of meters under low-illuminated and shade-illuminated environment. The experimental results indicates the superiority of the proposed numeric character recognition method.

• The Journal of the Korea Contents Association
• /
• v.22 no.6
• /
• pp.104-115
• /
• 2022

Recent companies that have grown to be the world's largest in the information technology industry and have a substantial impact on the global economy started as start-ups built on the founder's entrepreneurship. As a consequence, interest in entrepreneurship education has increased significantly. This research examined the essential elements and educational activities of the educational process among educators who undertake entrepreneurship-related education, using AHP methodologies. The study revealed that both educators rated competency education as more essential than character education. Within the character education dimension, habits were deemed more significant than attitudes, while within the competency education dimension, the technique education module was seen more important than the knowledge education module. This research is expected to contribute to the sustained expansion of domestic entrepreneurship education as well as to the response to opportunity entrepreneurship.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.1-25
• /
• 2020

In this paper, we suggest an application system architecture which provides accurate, fast and efficient automatic gasometer reading function. The system captures gasometer image using mobile device camera, transmits the image to a cloud server on top of private LTE network, and analyzes the image to extract character information of device ID and gas usage amount by selective optical character recognition based on deep learning technology. In general, there are many types of character in an image and optical character recognition technology extracts all character information in an image. But some applications need to ignore non-of-interest types of character and only have to focus on some specific types of characters. For an example of the application, automatic gasometer reading system only need to extract device ID and gas usage amount character information from gasometer images to send bill to users. Non-of-interest character strings, such as device type, manufacturer, manufacturing date, specification and etc., are not valuable information to the application. Thus, the application have to analyze point of interest region and specific types of characters to extract valuable information only. We adopted CNN (Convolutional Neural Network) based object detection and CRNN (Convolutional Recurrent Neural Network) technology for selective optical character recognition which only analyze point of interest region for selective character information extraction. We build up 3 neural networks for the application system. The first is a convolutional neural network which detects point of interest region of gas usage amount and device ID information character strings, the second is another convolutional neural network which transforms spatial information of point of interest region to spatial sequential feature vectors, and the third is bi-directional long short term memory network which converts spatial sequential information to character strings using time-series analysis mapping from feature vectors to character strings. In this research, point of interest character strings are device ID and gas usage amount. Device ID consists of 12 arabic character strings and gas usage amount consists of 4 ~ 5 arabic character strings. All system components are implemented in Amazon Web Service Cloud with Intel Zeon E5-2686 v4 CPU and NVidia TESLA V100 GPU. The system architecture adopts master-lave processing structure for efficient and fast parallel processing coping with about 700,000 requests per day. Mobile device captures gasometer image and transmits to master process in AWS cloud. Master process runs on Intel Zeon CPU and pushes reading request from mobile device to an input queue with FIFO (First In First Out) structure. Slave process consists of 3 types of deep neural networks which conduct character recognition process and runs on NVidia GPU module. Slave process is always polling the input queue to get recognition request. If there are some requests from master process in the input queue, slave process converts the image in the input queue to device ID character string, gas usage amount character string and position information of the strings, returns the information to output queue, and switch to idle mode to poll the input queue. Master process gets final information form the output queue and delivers the information to the mobile device. We used total 27,120 gasometer images for training, validation and testing of 3 types of deep neural network. 22,985 images were used for training and validation, 4,135 images were used for testing. We randomly splitted 22,985 images with 8:2 ratio for training and validation respectively for each training epoch. 4,135 test image were categorized into 5 types (Normal, noise, reflex, scale and slant). Normal data is clean image data, noise means image with noise signal, relfex means image with light reflection in gasometer region, scale means images with small object size due to long-distance capturing and slant means images which is not horizontally flat. Final character string recognition accuracies for device ID and gas usage amount of normal data are 0.960 and 0.864 respectively.

• Journal of KIISE:Software and Applications
• /
• v.34 no.1
• /
• pp.79-94
• /
• 2007

In the ubiquitous computing age which uses a high quantity network, mobile devices such as wearable and hand-held ones with a small tamers and a wireless communication module will be widely used in near future. Thus, a lot of researches about an augmented game on mobile devices have been attempted recently. The existing augmented games used a traditional 'backpack' system and a pattern marker. The 'backpack' system is expensive, cumbersome and inconvenient to use, and because of the pattern marker, it is only possible to play the game in the previously installed palace. In this paper, we propose an augmented game called Flying Cake using a face region to create the virtual object(character) without the pattern marker, which manually indicates an overlapped location of the virtual object in the real world, on a small and mobile PDA instead of the cumbersome hardware. Flying Cake is an augmented shooting game. This game supplies us with two types: 1) a single player which attacks a virtual character on images captured by a camera in an outdoor physical area, 2) dual players which attack the virtual character on images which we received through a wireless LAN. We overlap the virtual character on the face region using a face detection technique, and users play Flying Cake though attacking the virtual character. Flying Cake supplies new pleasure to flayers with a new game paradigm through an interaction between the user in the physical world captured by the PDA camera and the virtual character in a virtual world using the face detection.

• Journal of Information Technology Applications and Management
• /
• v.24 no.1
• /
• pp.33-43
• /
• 2017

The content curation service through big data analysis is receiving great attention in various content fields, such as film, game, music, and book. This service recommends personalized contents to the corresponding user based on user's preferences. The existing book curation systems recommended books to users by using bibliographic citation, user profile or user log data. However, these systems are difficult to recommend books related to character names or spatio-temporal information in text contents. Therefore, in this paper, we suggest a personalized book curation system based on integrated mining of a book. The proposed system consists of mining system, recommendation system, and visualization system. The mining system analyzes book text, user information or profile, and SNS data. The recommendation system recommends personalized books for users based on the analysed data in the mining system. This system can recommend related books using based on book keywords even if there is no user information like new customer. The visualization system visualizes book bibliographic information, mining data such as keyword, characters, character relations, and book recommendation results. In addition, this paper also includes the design and implementation of the proposed mining and recommendation module in the system. The proposed system is expected to broaden users' selection of books and encourage balanced consumption of book contents.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.10
• /
• pp.1291-1300
• /
• 2001

This article come to study the visualization representation technique of eye brain of person, basing on the ground of the dynamic graphics which is able to change the real time, manipulating the image as graphic factors of the multi-data. And the important thought in such realization is as follows ; corresponding the character points of human face and the parameter control value which obtains basing on the existing image recognition algorithm to the multi-dimensional data, synthesizing the image, it is to create the virtual image from the emotional expression according to the changing contraction expression. The proposed DyFIG system is realized that it as the completing module and we suggest the module of human face graphics which is able to express the emotional expression by manipulating and experimenting, resulting in realizing the emotional data expression description and technology.

• Journal of Korea Game Society
• /
• v.6 no.4
• /
• pp.79-87
• /
• 2006

This paper proposes the design and implementation of engine to control characters by using machine teaming techniques. Because the proposed engine uses the context data in the rum time as the knowledge data, there is a merit which the player can not easily recognize the behavior pattern of the intelligent character. To do this, the paper proposes to develop the module which gathers and trains the context data and the module which tests to decide the optimal context control for the given context data. The developed engine is ported to FEAR and run with Quake2 and experimented far the correctness of the development and its efficiency. The experiments show that the developed engine is operated well and efficiently within the limited time.

• Journal of Broadcast Engineering
• /
• v.25 no.5
• /
• pp.776-788
• /
• 2020

In this paper, a vehicle type recognition system using deep learning and a license plate recognition system are proposed. In the existing system, the number plate area extraction through image processing and the character recognition method using DNN were used. These systems have the problem of declining recognition rates as the environment changes. Therefore, the proposed system used the one-stage object detection method YOLO v3, focusing on real-time detection and decreasing accuracy due to environmental changes, enabling real-time vehicle type and license plate character recognition with one RGB camera. Training data consists of actual data for vehicle type recognition and license plate area detection, and synthetic data for license plate character recognition. The accuracy of each module was 96.39% for detection of car model, 99.94% for detection of license plates, and 79.06% for recognition of license plates. In addition, accuracy was measured using YOLO v3 tiny, a lightweight network of YOLO v3.

• Journal of the Korea Society for Simulation
• /
• v.16 no.1
• /
• pp.31-37
• /
• 2007

The importance of NPC's role in computer games is increasing. An NPC must perform its tasks by perceiving obstacles and other characters and by moving through them. It has been proposed to plan a natural-looking path against fixed obstacles by using a generalized visibility graph. In this paper we develop the execution module for an NPC to move efficiently along the path planned on the generalized visibility graph. The planned path consists of line segments and arc segments, so we define steering behaviors such as linear behaviors, circular behaviors, and an arriving behavior for NPC's movements to be realistic and utilize them during execution. The execution module also includes the collision detection capability to be able to detect dynamic obstacles and uses a decision tree to react differently according to the detected obstacles. The execution module is tested through the simulation based on the example scenario in which an NPC interferes the other moving NPC.