• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.10 no.3
• /
• pp.249-255
• /
• 2017

The finger vein recognition technology is a method to acquire a finger vein image by illuminating infrared light to the finger and to authenticate a person through processes such as feature extraction and matching. In order to recognize a finger vein, a 2D mask-based two-dimensional convolution method can be used to detect a finger edge but it takes too much computation time when it is applied to a low cost micro-processor or micro-controller. To solve this problem and improve the recognition rate, this study proposed an extraction method for the region of interest based on virtual core points and moving average filtering based on the threshold and absolute value of difference between pixels without using 2D convolution and 2D masks. To evaluate the performance of the proposed method, 600 finger vein images were used to compare the edge extraction speed and accuracy of ROI extraction between the proposed method and existing methods. The comparison result showed that a processing speed of the proposed method was at least twice faster than those of the existing methods and the accuracy of ROI extraction was 6% higher than those of the existing methods. From the results, the proposed method is expected to have high processing speed and high recognition rate when it is applied to inexpensive microprocessors.

• Journal of Platform Technology
• /
• v.6 no.1
• /
• pp.3-8
• /
• 2018

This paper proposes a technique for recognizing online handwritten cursive data obtained by tracing a motion trajectory while a user is in the 3D space based on a convolution neural network (CNN) algorithm. There is a difficulty in recognizing the virtual character input by the user in the 3D space because it includes both the character stroke and the movement stroke. In this paper, we divide syllable into consonant and vowel units by using labeling technique in addition to the result of localizing letter stroke and movement stroke in the previous study. The coordinate information of the separated consonants and vowels are converted into image data, and Korean handwriting recognition was performed using a convolutional neural network. After learning the neural network using 1,680 syllables written by five hand writers, the accuracy is calculated by using the new hand writers who did not participate in the writing of training data. The accuracy of phoneme-based recognition is 98.9% based on convolutional neural network. The proposed method has the advantage of drastically reducing learning data compared to syllable-based learning.

• Progress in Medical Physics
• /
• v.30 no.4
• /
• pp.120-127
• /
• 2019

Purpose: This study was designed to evaluate the dosimetric performance of Mobius3D by comparison with an aSi-based electronic portal imaging device (EPID) and Octavius 4D, which are conventionally used for patient-specific prescription dose verification. Methods: The study was conducted using nine patients who were treated by volumetric modulated arc therapy. To evaluate the feasibility of Mobius3D for prescription dose verification, we compared the QA results of Mobius3D to an aSi-based EPID and the Octavius 4D dose verification methods. The first was the comparison of the Mobius3D verification phantom dose, and the second was to gamma index analysis. Results: The percentage differences between the calculated point dose and measurements from a PTW31010 ion chamber were 1.6%±1.3%, 2.0%±0.8%, and 1.2%±1.2%, using collapsed cone convolution, an analytical anisotropic algorithm, and the AcurosXB algorithm respectively. The average difference was found to be 1.6%±0.3%. Additionally, in the case of using the PTW31014 ion chamber, the corresponding results were 2.0%±1.4%, 2.4%±2.1%, and 1.6%±2.5%, showing an average agreement within 2.0%±0.3%. Considering all the criteria, the Mobius3D result showed that the percentage dose difference from the EPID was within 0.46%±0.34% on average, and the percentage dose difference from Octavius 4D was within 3.14%±2.85% on average. Conclusions: We conclude that Mobius3D can be used interchangeably with phantom-based dosimetry systems, which are commonly used as patient-specific prescription dose verification tools, especially under the conditions of 3%/3 mm and 95% pass rate.

• Journal of the Semiconductor & Display Technology
• /
• v.21 no.2
• /
• pp.68-73
• /
• 2022

In this study, we produced the surface defect detection model to automatically detect defect bone plates using a deep learning algorithm. Bone plates with a width and a length of 50 mm are most used for fracture treatment. Normal bone plates and defective bone plates were printed on the 3d printer. Normal bone plates and defective bone plates were photographed with 1,080 pixels using the webcam. The total quantity of collected images was 500. 300 images were used to learn the defect detection model. 200 images were used to test the defect detection model. The mAP(Mean Average Precision) method was used to evaluate the performance of the surface defect detection model. As the result of confirming the performance of the surface defect detection model, the detection accuracy was 96.3 %.

• Journal of Korea Multimedia Society
• /
• v.22 no.6
• /
• pp.655-664
• /
• 2019

In this paper, we propose a novel way of combining multiple deep convolutional neural network (DCNN) architectures which work well for accurate video face identification by adopting a serial combination of 3D and 2D DCNNs. The proposed method first divides an input video sequence (to be recognized) into a number of sub-video sequences. The resulting sub-video sequences are used as input to the 3D DCNN so as to obtain the class-confidence scores for a given input video sequence by considering both temporal and spatial face feature characteristics of input video sequence. The class-confidence scores obtained from corresponding sub-video sequences is combined by forming our proposed class-confidence matrix. The resulting class-confidence matrix is then used as an input for learning 2D DCNN learning which is serially linked to 3D DCNN. Finally, fine-tuned, serially combined DCNN framework is applied for recognizing the identity present in a given test video sequence. To verify the effectiveness of our proposed method, extensive and comparative experiments have been conducted to evaluate our method on COX face databases with their standard face identification protocols. Experimental results showed that our method can achieve better or comparable identification rate compared to other state-of-the-art video FR methods.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.2
• /
• pp.175-180
• /
• 2002

In this paper, the system performace with the convolution rode using a Viterbi decoding and the one tap LMS(Least Meam Square) equalizer applied to the OFDM(Orthogonal Frequency Division Multiplexing) system, is analyzed through computer simulation. DMRS(Digital Microwave Radio System)is modeled as Rummler fading channel. In Simulation result, we known that the coding system improved about 3.6dB~10.5dB when BER is 10 $^3$and b is 0.1~0.2 in case of 16QAM(Qurdrature Amplitude Modulation). Also, we known that was improved about 19.7dB when the b is 0.1 and was demanded about 10.5dB when the b is 0.2 in case of 64QAM. we known that the soft decision improved about 2~0.9dB when the b is 0.1~0.2 in case of 16QAM and about 3.3~7.8dB in case of 64QAM. In the equalizer system, efficiency improved from the case of that Eb/No is more than 13dB.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.3
• /
• pp.1076-1094
• /
• 2022

Technology for emotion recognition is an essential part of human personality analysis. To define human personality characteristics, the existing method used the survey method. However, there are many cases where communication cannot make without considering emotions. Hence, emotional recognition technology is an essential element for communication but has also been adopted in many other fields. A person's emotions are revealed in various ways, typically including facial, speech, and biometric responses. Therefore, various methods can recognize emotions, e.g., images, voice signals, and physiological signals. Physiological signals are measured with biological sensors and analyzed to identify emotions. This study employed two sensor types. First, the existing method, the binary arousal-valence method, was subdivided into four levels to classify emotions in more detail. Then, based on the current techniques classified as High/Low, the model was further subdivided into multi-levels. Finally, signal characteristics were extracted using a 1-D Convolution Neural Network (CNN) and classified sixteen feelings. Although CNN was used to learn images in 2D, sensor data in 1D was used as the input in this paper. Finally, the proposed emotional recognition system was evaluated by measuring actual sensors.

• Journal of The Korean Association of Information Education
• /
• v.15 no.3
• /
• pp.517-524
• /
• 2011

This paper presents a fast blind deconvolution method that produces a deblurring result from a single image in only a few seconds. The high speed of our method is enabled by considering the Discrete Fourier Transform (DFT), and its relation to filtering and convolution, and fast computation of Moore-Penrose inverse matrix. How can we predict the behavior of an arbitrary filter, or even more to the point design a filter to achieve certain specifications. The idea is to study the frequency response of the filter. This concept leads to an useful convolution formula. A Matlab implementation of our method usually takes less than one minute to deblur an image of moderate size, while the deblurring quality is comparable.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.5
• /
• pp.1778-1797
• /
• 2021

Document image binarization is an important pre-processing step in document analysis and archiving. The state-of-the-art models for document image binarization are variants of encoder-decoder architectures, such as FCN (fully convolutional network) and U-Net. Despite their success, they still suffer from three limitations: (1) reduced feature map resolution due to consecutive strided pooling or convolutions, (2) multiple scales of target objects, and (3) reduced localization accuracy due to the built-in invariance of deep convolutional neural networks (DCNNs). To overcome these three challenges, we propose an improved semantic segmentation model, referred to as DP-LinkNet, which adopts the D-LinkNet architecture as its backbone, with the proposed hybrid dilated convolution (HDC) and spatial pyramid pooling (SPP) modules between the encoder and the decoder. Extensive experiments are conducted on recent document image binarization competition (DIBCO) and handwritten document image binarization competition (H-DIBCO) benchmark datasets. Results show that our proposed DP-LinkNet outperforms other state-of-the-art techniques by a large margin. Our implementation and the pre-trained models are available at https://github.com/beargolden/DP-LinkNet.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.11a
• /
• pp.15-18
• /
• 2002

본 논문에서는 다양한 채널을 가진 음악 과일에 대하여 입체 음향 효과를 줄 수 있는 시스템을 개발 하였다. 그러기 위하여 3D 사운드 기술 중에 가장 대표적으로 알려진 HRTF(머리전달 함수)를 원음에 콘볼루션(Convolution)하는 방식으로 음상정위 모듈을 구현하였으며 음장감을 부여하기 위해 잔향 효과(Reverberation)효과를 추가하고 크로스토크 현상 제거를 위해 트랜스오럴(Transaural) 필터를 추가하였다. 이런 입체음향 기술을 가지고 여러 채널을 가진 음악 파일에 적용시켜서 다채널 입체음향 효과를 낸 수 있는 시뮬레이터를 구현해 보았다. 시스템 구현에는 한정된 채널이 아닌 다양한 채널에 대한 효과를 낼 수 있도록 하였으며 기본적인 실험으로는 미디를 바탕으로한 5개의 채널에 대하여 실험하여 이를 증명해 보았다.