• Journal of Information Processing Systems
• /
• v.17 no.2
• /
• pp.411-425
• /
• 2021

In this paper, a vehicle recognition algorithm based on deep convolutional neural network and compression dictionary is proposed. Firstly, the network structure of fine vehicle recognition based on convolutional neural network is introduced. Then, a vehicle recognition system based on multi-scale pyramid convolutional neural network is constructed. The contribution of different networks to the recognition results is adjusted by the adaptive fusion method that adjusts the network according to the recognition accuracy of a single network. The proportion of output in the network output of the entire multiscale network. Then, the compressed dictionary learning and the data dimension reduction are carried out using the effective block structure method combined with very sparse random projection matrix, which solves the computational complexity caused by high-dimensional features and shortens the dictionary learning time. Finally, the sparse representation classification method is used to realize vehicle type recognition. The experimental results show that the detection effect of the proposed algorithm is stable in sunny, cloudy and rainy weather, and it has strong adaptability to typical application scenarios such as occlusion and blurring, with an average recognition rate of more than 95%.

• Journal of Korea Multimedia Society
• /
• v.8 no.2
• /
• pp.225-232
• /
• 2005

A PDA has limitation of storage space and inconvenience of Pen-based equipment to input. To solve these problems, this paper implemented that previous Speech Recognition provides using XML(extensile Markup Language) Web Service and PDA Client is possible to use speech command by request Speech Recognition Web Service. This system consists of Mobile Client and Speech Recognition Web Service Provider. The Mobile Client transmits user information and speech data to the Speech Recognition Web Service Provider using DIME(Direct Internet Message Encapsulation). The Speech Recognition Web Service Provider returns recognition results and user name only to an authorized Mobile Client. So We are able to use speech service without installing Speech Recognition or Speech DB in PDA.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.8 no.3
• /
• pp.179-184
• /
• 2013

A voice recognition technology as a technology fundament plays an important role in medical devices with smart functions. This paper describes the implementation of a control system that can be utilized as a part of illumination equipment for medical applications (IEMA) based on a voice recognition. The control system can essentially be divided into five parts, the microphone, training part, recognition part, memory part, and control part. The system was implemented using the RSC-4x evaluation board which is included the micro-controller for voice recognition. To investigate the usefulness of the implemented control system, the experiments of the recognition rate was carried out according to the input distance for voice recognition. As a result, the recognition rate of the control system was more than 95% within a distance between 0.5 and 2m. The result verified that the implemented control system performs well as the smart control system based for an IEMA.

• Journal of Information Processing Systems
• /
• v.16 no.1
• /
• pp.224-237
• /
• 2020

Recognition of banknote serial number is one of the important functions for intelligent banknote counter implementation and can be used for various purposes. However, the previous character recognition method is limited to use due to the font type of the banknote serial number, the variation problem by the solid status, and the recognition speed issue. In this paper, we propose an aspect ratio based character region segmentation and a convolutional neural network (CNN) based banknote serial number recognition method. In order to detect the character region, the character area is determined based on the aspect ratio of each character in the serial number candidate area after the banknote area detection and de-skewing process is performed. Then, we designed and compared four types of CNN models and determined the best model for serial number recognition. Experimental results showed that the recognition accuracy of each character was 99.85%. In addition, it was confirmed that the recognition performance is improved as a result of performing data augmentation. The banknote used in the experiment is Indian rupee, which is badly soiled and the font of characters is unusual, therefore it can be regarded to have good performance. Recognition speed was also enough to run in real time on a device that counts 800 banknotes per minute.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.10
• /
• pp.217-223
• /
• 2009

Vocabulary recognition system of Mobile terminal is executed statistical method for vocabulary recognition and used statistical grammar recognition system using N-gram. If limit arithmetic processing capacity in memory of vocabulary to grow then vocabulary recognition algorithm complicated and need a large scale search space and many processing time on account of impossible to process. This study suggest vocabulary recognition optimize using MLHF System. MLHF separate acoustic search and lexical search system using FLaVoR. Acoustic search feature vector of speech signal extract using HMM, lexical search recognition execution using Levenshtein distance algorithm. System performance as a result of represent vocabulary dependence recognition rate of 98.63%, vocabulary independence recognition rate of 97.91%, represent recognition speed of 1.61 second.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.5 s.37
• /
• pp.353-362
• /
• 2005

Recently, study of Bimodal recognition has become very active. In this paper we propose a Bimodal user recognition system that uses face information and audio information. Face recognition consists of face detection step and face recognition step. Face detection uses AdaBoost to find face candidate area. After finding face candidates, PCA feature extraction is applied to decrease the dimension of feature vector. And then, SVM classifiers are used to detect and recognize face. Audio recognition uses MFCC for audio feature extraction and HMM is used for audio recognition. Experimental results show that the Bimodal recognition can improve the user recognition rate much more than audio only recognition, especially in the Presence of noise.

• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.313-316
• /
• 2002

The need to protect a main information and control a personal access using Biometrics in the intelligent building system is extended with growing the information and communication technologies in recent years. The fingerprint recognition technology is one of the biometrics methods available that has been widely used in various applications. But the present architecture of fingerprint recognition system has many of the problems because of centralized control network architecture. But we can solve the problems with distributed control network architecture. In this paper, the fingerprint recognition system uses the merit of the LonWorks. it is able to easily modularize the system and make up for the weak points in the typical fingerprint recognition system.

• Journal of the Korea Fashion and Costume Design Association
• /
• v.16 no.1
• /
• pp.77-89
• /
• 2014

This study looked into the effect of the life style of single females in 20s and 30s on beauty behavior recognition, and spss 17.0 is used for data analysis method. As for the statistical analysis method in order to validate the measurement tools, reliability verification is conducted and life style groups are sampled using K-means taking into account factor scores by life style. To find out the difference between general beauty behavior recognition and life style, descriptive statistics and One Way ANOVA were carried out, and Duncan Test was implemented for the post examination method. Multiple regression analysis was also carried out to figure out the effect of life style on beauty behavior recognition. The result is as follows. First, according to the results of reliability verification and factor analysis for the lifestyle type and the recognition of the behavior for beauty, the types of the life style of the subjects were divided into Economic Utility, Convention Conservatism, Self Development, Showy Consumption, and Appearance Oriented, and the recognition of the behavior for beauty was named as Makeup and Hair, Cosmetic Surgery, Body Care, and Skin Care. Second, as to the recognition of the behavior for beauty based upon the lifestyle, the Appearance Oriented in Showy Consumption recorded the highest. Third, the analysis of the influence of the style on the recognition of the behavior for beauty showed that the behavior recognition for Makeup and Hair and for Skin Care was affected by the life style of Self Development, Showy Consumption, and Appearance Oriented; the behavior recognition for Cosmetic Surgery was affected by the life style of Conventional Conservatism, Showy Consumption, and Appearance Oriented; and again the behavior recognition for Body Care was by that of Economical Utility and Showy Consumption.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1140-1145
• /
• 2003

Image normalization is one of the important areas in pattern recognition. Also, log-polar images are useful in the sense that their image data size is reduced dramatically comparing with conventional images and it is possible to develop faster pattern recognition algorithms. Especially, the log-polar image is very similar with the structure of human eyes. However, there are almost no researches on pattern recognition using the log-polar images while a number of researches on visual tracking have been executed. We propose an image normalization technique of log-polar images using momentums applicable for affine-invariant pattern recognition. We handle basic distortions of an image including translation, rotation, scaling, and skew of a log-polar image. The algorithm is experimented in a PC-based real-time vision system successfully.

• Journal of The Korea Institute of Healthcare Architecture
• /
• v.27 no.1
• /
• pp.17-34
• /
• 2021

Purpose: The purpose of this study was to find out the difference in the priority level of recognition for universal design application in public spaces by gender. Method: ANOVA(analysis of variance) and post-hoc test were conducted to determine the priority level of recognition and pattern for the disabled, non-disabled, and experts classified into males and females. Results: There was no gender difference in the comparison by sector for all males and females. However, in comparing of domains and facilities, women showed a higher level of recognition than men in the building sector and cross domain. When comparing space consumers and producers by dividing them into male and female groups, women showed a higher level of recognition than men in producers, but there was no gender difference between consumers. In comparison by sector, domain and facility, women producers also showed a higher level of recognition in the road sector, park and recreation sector, sidewalk domain, four-spaces in the park and recreation sector, and six-spaces in the building sector than men producers. Also, in the building sector, women producers and consumers showed a higher recognition level than men. Comparing the disabled, non-disabled people and experts by dividing them into male and female groups, in the case of non-disabled people and experts, women showed a higher level of recognition than men, while men showed a higher level of recognition than women in the disabled. In addition, there were differences in recognition patterns in many spaces and facilities by gender. Implications: This study is meaningful in comparing the differences in the priority level of recognition and patterns between men and women to apply universal design for people of all ages and both sexes.